diff --git a/.agents/skills/hyb-oa-delivery/SKILL.md b/.agents/skills/hyb-oa-delivery/SKILL.md
index 43ca391d..50cc9096 100644
--- a/.agents/skills/hyb-oa-delivery/SKILL.md
+++ b/.agents/skills/hyb-oa-delivery/SKILL.md
@@ -23,7 +23,7 @@ Read only what is relevant:
 - WROUTE2 locks the post-WFULLSMOKE1 route: first close out test-account evidence, then ask for WREMOTE1/WFULLSMOKE2 remote closeout and WEXT1-FORMAL formal-template outing+overtime manual test, then ask WREAL1 with exactly 1 real employee + 1 approver/admin observer, then prioritize `wecom_attendance_approval_mirror_v1` minimum complete mirror. WTCLOSE1 adds the small-team execution rule: because the company has fewer than 10 people, WAPPR1 targets option A for leave/missed-punch/outing/overtime; formal HR/attendance approvals use one direct manager or configured approval role with admin fallback/blocking when the approver chain is missing, unavailable, invisible, unbound, or self-approval. WRULE1 enforces this as an OA-side validator over archived WeCom approval details; `PASS_RULE_MATCH` or explainable `PASS_WITH_ADMIN_FALLBACK` can support test-account closeout, while `FAIL_APPLICANT_CAN_CHOOSE`, `FAIL_SELF_APPROVAL`, `FAIL_APPROVER_MISMATCH`, missing manual gates, or external config blockers stop WEXT1-FORMAL/WREAL1. WTCLOSE1 only authorizes WAUTO1/WFULLSMOKE1 aggregation and WEXT1LD dry-run; WTCLOSE2 only performs final test-account evidence closeout. Neither authorizes WeCom writes, real employees, outing/overtime fixture creation, emergency mode, writeback, or RuleOps.
 - WAPPR1 2026-05-21 manual gate result: the user set the test employee's direct manager to the test supervisor and confirmed on a real device that leave, missed punch, outing, and overtime all auto-fill the test supervisor and the applicant cannot delete/change/add approvers. Treat all four templates as A / `PASS_APPROVER_RULES_READY`, but still require WRULE1 actual-chain validation, WROAD2-BIZ1 business evidence, WTCLOSE2 final test-account closeout, and explicit WEXT1-FORMAL/WREAL1 question gates before any WeCom write or real-employee trial.
 - WTCLOSE2/WTCLOSE2F adds `scripts/qa/run-wecom-test-account-final-closeout.sh` as the final automatic test-account closeout before WEXT1-FORMAL. Enterprise WeChat API checks must run from the deployment server trusted IP, so use `bash scripts/qa/run-wecom-test-account-final-closeout.sh --remote`; local direct execution is blocked by default and may only be used for syntax/static gates. The remote run reruns `/ready`, WAUTO1, WFULLSMOKE1 with `WFULLSMOKE_ALLOW_PUBLIC_WRITES=0`, outing/overtime dry-run without live flags, and WRULE1 with `--manual-gate=all:A`; it must not set WeCom write switches, create templates/approvals, enable real employees, or treat WAPPR1=A as live approval. Do not treat previous local missing-env/DB WTCLOSE2 output as a WeCom business blocker.
-- WREMOTE1/WFULLSMOKE2/WEXT1-FORMAL supersedes the earlier fixture-creation route for the next step: first deploy/sync and run WTCLOSE2 from the deployment server trusted IP, then run WFULLSMOKE2 from the deployment server with `WFULLSMOKE_ALLOW_PUBLIC_WRITES=1` when OA `TEST-SMOKE-WFULL` public test data is approved, then use existing formal WeCom outing/overtime templates by having `SMOKE002` manually submit and `SMOKE101` review details before approving. Do not auto-create outing/overtime OA-POC templates or auto-submit approvals in this route. WREAL1 uses `/tmp/hyb-oa-wreal1.env` for real employee/approver values and must not write real employee identifiers to repo docs, Issue CSV, or chat.
+- WREMOTE1/WFULLSMOKE2/WEXT1-FORMAL supersedes the earlier fixture-creation route for the next step: first deploy/sync and run WTCLOSE2 from the deployment server trusted IP, then run WFULLSMOKE2 from the deployment server with `WFULLSMOKE_ALLOW_PUBLIC_WRITES=1` when OA `TEST-SMOKE-WFULL` public test data is approved, then use existing formal WeCom outing/overtime templates by having `SMOKE002` manually submit and `SMOKE101` review details before approving. WEXT1F adds `WEXT1F_MANUAL_CONFIRMATION=SMOKE002_SUBMITTED_SMOKE101_REVIEWED bash scripts/qa/run-wecom-formal-outing-overtime-evidence.sh --remote` for the post-manual evidence run: it explicitly pulls `--target=outing --target=overtime` with `--filter-creator-employee-no=SMOKE002`, may persist only to OA archive via `WECOM_ARCHIVE_WRITE=1` + `--persist-to-db`, and then runs archive offline read + WRULE1. Missing manual confirmation must produce `PAUSED_MANUAL_CONFIRMATION` and no WeCom API calls. Do not auto-create outing/overtime OA-POC templates, auto-submit approvals, write WeCom, broadly pull formal-template records from other employees, or run WeCom API evidence locally in this route. WREAL1 uses `/tmp/hyb-oa-wreal1.env` for real employee/approver values and must not write real employee identifiers to repo docs, Issue CSV, or chat.
 - WAPC2FLIVE supersedes the WAPC2F dry-run gap for OA-POC samples: after user approval, the runner was corrected to the official `template_name` + `property` controls create-template payload, deployed with `SYNC_ENV=0`, and the public live report `/app/artifacts/qa-flow/wecom-approval-fixtures-20260519T053400Z/summary.json` reached `PASS_FIXTURE_CREATED`; leave/missed-punch OA-POC templates and low-risk test approvals were created and retained for audit, with WAPC2 auto-rerun `PASS_APPROVAL_POC`. WCBK1 has added and deployed read-only `GET/POST /api/wecom/approval/callback`, `wecomCallbackCrypto`, `npm run check:wecom-approval-callback-poc`, and the `wecom-callback` process gate. After remote callback config was set, public report `/app/artifacts/qa-flow/wecom-approval-callback-poc-20260519T083422Z/summary.json` reached `PASS_CALLBACK_URL_VERIFIED`; after `SMOKE101` approved an existing OA-POC missed-punch test approval, `/app/artifacts/qa-flow/wecom-approval-callback-poc-20260519T092648Z/summary.json` reached `PASS_CALLBACK_EVENT_CONFIRMED` with one redacted event. WACP3/WACP3R has added and deployed `npm run check:wecom-approval-compensation-poc`, `POST /cgi-bin/oa/getapprovalinfo` read support, and automatic template discovery by `SMOKE002` creator; WREAD1 deploy reran WACP3R at `/app/artifacts/qa-flow/wecom-approval-compensation-poc-20260519T125515Z/summary.json` and then generated `/app/artifacts/qa-flow/wecom-approval-readability-gate-20260519T125532Z/summary.json` as `PAUSED_MANUAL_CONFIRMATION`. WREAD1F then fixed duplicate OA-POC template names by adding timestamp uniqueness and created fresh leave/missed-punch approvals at `/app/artifacts/qa-flow/wecom-approval-fixtures-20260519T131958Z/summary.json`; final WREAD1 manual report `/app/artifacts/qa-flow/wecom-approval-readability-gate-20260519T133628Z/summary.json` is `PASS_READABLE`. This clears leave/missed-punch readability only; mirror data model, minimum archive, emergency mode, writeback, RuleOps, and outing/overtime sample decisions remain later gates.
 - WEGOV1 changes the current WeCom route to "Enterprise WeChat owns mature HR/attendance flows; OA stores self-owned history, audit, entry, messages, and emergency fallback." Treat four domains explicitly: WeCom-owned domain is check-in, leave, missed punch, outing, overtime, leave balance, and HR assistant; OA-owned domain remains reimbursement, procurement, contracts, projects, payments, invoices, bank receipts, sensitive attachments, and operating audit; OA archive domain stores WeCom business history, full external keys, and necessary snapshots; OA emergency domain is manually enabled only when WeCom is unavailable. Do not assume a full isolated WeCom sandbox; use the controlled test domain `SMOKE002 -> SMOKE101`, `OA-POC` templates, low-risk approvals, synthetic tests, public live runners, and limited real-device confirmation. WARCH1 now supplies the minimum leave/missed-punch archive implementation, WARCH2 supplies the formal read-only history query surface, and WROAD2 supplies controlled ops daily checks. Do not start deep mirror until WROAD2-CLOSEOUT, WROAD2-BIZ1, WAPPR1, WRULE1, WAUTO1/WEXT1LD, WFULLSMOKE1, WTCLOSE2, WEXT1-FORMAL, WREAL1 question gates or explicit approval, and the mirror stage's own contract are complete.
 - For reimbursement, invoice, OCR, duplicate invoice, or attachment work, read `docs/报销发票附件规则.md` and `references/reimbursement.md`.
diff --git a/docs/trialrun/人工试运行报告.md b/docs/trialrun/人工试运行报告.md
index 7f099f8f..99a4cfa1 100644
--- a/docs/trialrun/人工试运行报告.md
+++ b/docs/trialrun/人工试运行报告.md
@@ -580,6 +580,18 @@ Go 结论边界：
 | 外出/加班 | 路线修正 | 不自动创建 OA-POC 外出/加班模板；由 `SMOKE002` 在正式模板真机提交，`SMOKE101` 先看详情再审批，系统只读补偿、归档和 WRULE1。 |
 | 真实试用 | 待问答 | 首批仍为 1 名真实员工 + 1 名审批人/管理员观察；真实员工号只通过 `/tmp/hyb-oa-wreal1.env` 临时文件提供，不进仓库。 |
 
+## WEXT1F 正式模板外出/加班证据脚本
+
+执行时间：2026-05-22 建档并实现。本轮承接用户确认“可以”进入正式模板外出/加班测试准备，但系统仍不自动写企业微信、不创建模板、不自动提交审批单。
+
+| 项目 | 状态 | 证据与边界 |
+| --- | --- | --- |
+| 执行合同 | 已建档 | 新增 `issues/2026-05-22_14-23-19-wecom-formal-outing-overtime-evidence-v1.csv`，ID 为 `WEXT1F-0..WEXT1F-7`。 |
+| 只读证据脚本 | 已新增 | 新增 `scripts/qa/run-wecom-formal-outing-overtime-evidence.sh`，真实业务运行必须使用 `--remote` 从部署服务器 server 容器执行；本地直接执行默认拒绝，即使伪造远端环境变量也会因非 `/app` 容器路径拒绝。 |
+| 人工前置 | 待执行 | `SMOKE002` 需在企业微信正式外出和加班模板提交低风险测试单，并确认审批人自动带出测试主管且不能改人；`SMOKE101` 需先打开详情页再同意或驳回。 |
+| 系统校验 | 待人工后远端执行 | 脚本必须带 `WEXT1F_MANUAL_CONFIRMATION=SMOKE002_SUBMITTED_SMOKE101_REVIEWED`，否则只输出 `PAUSED_MANUAL_CONFIRMATION`，不跑企微 API；正式拉取显式使用 `--target=outing --target=overtime` 和 `--filter-creator-employee-no=SMOKE002`，只收 `SMOKE002` 相关单据，在 `WECOM_ARCHIVE_WRITE=1` + `--persist-to-db` 双开关下写 OA 自建留存，然后执行外出/加班离线归档读取、WRULE1 审批链校验和隐私扫描。 |
+| 写入边界 | 已固定 | 不设置 `WECOM_APPROVAL_FIXTURE_WRITE=1`，不调用 create_template/applyevent，不创建 OA-POC 外出/加班样件，不启用真实员工；报告不得泄露完整 `sp_no`、template_id、userid、手机号、邮箱、token、Secret、EncodingAESKey、表单值、附件、评论或原始 payload。 |
+
 ## 历史恢复
 
 需要追溯 R70-R91 的详细流水时，使用：
diff --git a/docs/企微优先开发路线图.md b/docs/企微优先开发路线图.md
index 8b2de1ad..c7560719 100644
--- a/docs/企微优先开发路线图.md
+++ b/docs/企微优先开发路线图.md
@@ -2,8 +2,8 @@
 
 **文档状态**：当前开发路线图事实源。  
 **分类**：开发治理 / 路线图。  
-**适用时间**：WDRG / WCP1 / WHMSLIVE / WOUA / WAPI / WRRA / WOUA-FIX / WRV1 / WRA1 / WMSG1 / WAPC1 / WAPC2 / WPROC1 / WPROC2 / WAPC2F / WAPC2FLIVE / WCBK1 / WEGOV1 / WACP3 / WACP3R / WREAD1 / WREAD1F / WADM1 / WARCH1 / WARCH2 / WROAD2 / WARCH3 / WCHECK1 / WEXT1 / WOPS1 / WROAD2-CLOSEOUT / WROAD2-BIZ1 / WAPPR1 / WAUTO1 / WEXT1LD / WFULLSMOKE1 / WROUTE2 / WTCLOSE1 / WRULE1 / WTCLOSE2 / WTCLOSE2F / WREMOTE1 / WFULLSMOKE2 / WEXT1-FORMAL / WREAL1-PREP 路线选择之后。
-**关联 Issue**：`WDRG,WCP1,WHMS,WHMSLIVE,WOUA,WAPI,WRRA,WOUAFIX,WRV1,WRA1,WMSG1,WAPC1,WAPC2,WPROC1,WPROC2,WAPC2F,WAPC2FLIVE,WCBK1,WEGOV1,WACP3,WACP3R,WREAD1,WREAD1F,WADM1,WARCH1,WARCH2,WROAD2,WARCH3,WCHECK1,WEXT1,WOPS1,WROAD2C,WROAD2B,WAPPR1,WAUTO1,WEXT1LD,WFULLSMOKE1,WROUTE2,WTCLOSE1,WRULE1,WTCLOSE2,WTCLOSE2F,WREMOTE1,WFULLSMOKE2,WEXT1-FORMAL,WREAL1P`。
+**适用时间**：WDRG / WCP1 / WHMSLIVE / WOUA / WAPI / WRRA / WOUA-FIX / WRV1 / WRA1 / WMSG1 / WAPC1 / WAPC2 / WPROC1 / WPROC2 / WAPC2F / WAPC2FLIVE / WCBK1 / WEGOV1 / WACP3 / WACP3R / WREAD1 / WREAD1F / WADM1 / WARCH1 / WARCH2 / WROAD2 / WARCH3 / WCHECK1 / WEXT1 / WOPS1 / WROAD2-CLOSEOUT / WROAD2-BIZ1 / WAPPR1 / WAUTO1 / WEXT1LD / WFULLSMOKE1 / WROUTE2 / WTCLOSE1 / WRULE1 / WTCLOSE2 / WTCLOSE2F / WREMOTE1 / WFULLSMOKE2 / WEXT1-FORMAL / WEXT1F / WREAL1-PREP 路线选择之后。
+**关联 Issue**：`WDRG,WCP1,WHMS,WHMSLIVE,WOUA,WAPI,WRRA,WOUAFIX,WRV1,WRA1,WMSG1,WAPC1,WAPC2,WPROC1,WPROC2,WAPC2F,WAPC2FLIVE,WCBK1,WEGOV1,WACP3,WACP3R,WREAD1,WREAD1F,WADM1,WARCH1,WARCH2,WROAD2,WARCH3,WCHECK1,WEXT1,WOPS1,WROAD2C,WROAD2B,WAPPR1,WAUTO1,WEXT1LD,WFULLSMOKE1,WROUTE2,WTCLOSE1,WRULE1,WTCLOSE2,WTCLOSE2F,WREMOTE1,WFULLSMOKE2,WEXT1-FORMAL,WEXT1F,WREAL1P`。
 **最近审计**：2026-05-22。
 **不得误读**：本文只控制后续开发节奏、阶段顺序和门禁。WHMSLIVE 后只代表企微 HR 主数据镜像已部署并受控验证；WAPI 后进一步固定企微 API/Secret 口径并转为运行优先；WRRA 后明确 `PASS_WITH_FINDINGS` 只代表只读对账成功且存在差异，不代表企微考勤/审批镜像、OA 应急接管、恢复回写或 RuleOps 已完成运行时实现。WMSG1 只证明企业微信应用消息可达和 Deeplink 可回到 OA；WAPC1 只固化审批配置准备清单；WAPC2 只新增请假/补卡模板详情和低风险 `sp_no` 审批详情只读 POC，不代表真实回调或镜像能力已完成。WPROC1 后，WAPC2 不直接放行深度镜像，必须先完成自动样件、回调、补偿拉取和镜像数据模型设计门禁。WPROC2 只新增可执行 process gate。WAPC2FLIVE 已创建 OA-POC 请假/补卡测试模板和低风险测试审批单，并自动复跑 WAPC2 得到 `PASS_APPROVAL_POC`。WCBK1 已完成真实回调 URL 验证和一条现有 OA-POC 测试审批单事件确认，公网报告为 `PASS_CALLBACK_EVENT_CONFIRMED`；WEGOV1 后路线明确为“企微成熟功能优先上线，OA 做自建留存、审计、入口、消息和应急兜底”。WACP3R 已将补偿拉取复跑修正为自动按 `SMOKE002` creator 发现 OA-POC 请假/补卡模板，公网报告 `/app/artifacts/qa-flow/wecom-approval-compensation-poc-20260519T115944Z/summary.json` 为 `PASS_WITH_FINDINGS`。WREAD1F 新样例后，审批人真机确认请假和补卡消息卡片、列表、详情页均可看清；远端人工结果报告 `/app/artifacts/qa-flow/wecom-approval-readability-gate-20260519T133628Z/summary.json` 为 `PASS_READABLE`。但同意按钮单击即通过仍是操作风险，需在 WARCH1/WARCH2 的留存、审计、提示或操作规程中继续处理。WADM1 只完成自建留存/镜像数据模型设计；WARCH1 只实现请假/补卡最小审批历史留存、回调事件入库、补偿拉取入库和 QA-only 离线读取证明；WARCH2 只提供 OA DB 只读查询 API 和 PC/移动历史页，不调用企微 API、不读取企微 Secret、不代表完整镜像、外出/加班、应急接管、恢复回写或 RuleOps 已完成。WROAD2 后当前策略是测试账号受控上线和每日脚本化巡检，允许补偿归档写 OA 自建留存库，但不写企业微信、不扩大真实员工范围；WARCH3 将每日巡检升级为服务器 `systemd timer` 与脱敏企微告警，WCHECK1 增加打卡规范化明细、日/月摘要和规则快照留存，WEXT1 只完成外出/加班 dry-run 支持。WOPS1 已公网验证巡检告警降噪：`HYD000006` 是管理员 `wangshunli` 的 employee_no，脚本不再把姓名字符串当员工号；已知 finding 不重复发送，只有阻断、脚本失败、新增 finding 或状态恶化才告警。WROAD2-BIZ1 在三天内补充测试账号人工业务冒烟；WAPPR1 将申请人任意选择审批人列为真实试用前阻断 finding。WAUTO1 只做只读证据汇总，WEXT1LD 只做外出/加班 live 前 dry-run，均不授权 live 创建。WFULLSMOKE1 是三天全域最终目标测试总控，只聚合企微 HR/考勤、OA 主办域、移动端和运维巡检证据；它允许在公网试运行环境创建带 `TEST-SMOKE-WFULL` 标记的 OA 测试业务数据，但不写企业微信、不创建外出/加班样件、不做真实付款、不接真实支付、不写 ERP、不导入银行流水、不启用真实员工。WREAL1-PREP 只准备真实小范围试用清单，不启用真实员工；后续顺序收紧为 WROAD2-CLOSEOUT 三天验收 -> WROAD2-BIZ1 三天业务冒烟 -> WAPPR1 审批规则规范化 -> WAUTO1 自动证据汇总 -> WEXT1LD 外出/加班 live 前 dry-run -> WFULLSMOKE1 三天全域最终目标测试 -> WEXT1-FORMAL 正式模板外出/加班测试 -> 请假/补卡/外出/加班四类测试号证据和审批规则门禁通过后再问真实小范围试用。上述阶段仍不代表完整深度镜像、应急接管、恢复回写或 RuleOps 已完成。
 **WTCLOSE2F 顺序修正**：上一条长段中的“后续顺序”以本条为准：WROAD2-CLOSEOUT -> WROAD2-BIZ1 -> WAPPR1 -> WRULE1 -> WAUTO1 -> WEXT1LD -> WFULLSMOKE1 -> WTCLOSE2/WTCLOSE2F 远端可信 IP 收口 -> WREMOTE1/WFULLSMOKE2 远端收口 -> WEXT1-FORMAL 正式模板人工提交问答 -> WREAL1 问答 -> 最小完整镜像。
@@ -13,6 +13,7 @@
 **WRULE1 执行口径**：采用“单一组织规则，两个执行域”。OA 维护员工、直属主管、业务角色和管理员兜底事实；企业微信只执行 HR/考勤审批；OA 通过 `check:wecom-approval-rule-validator` 读取已归档企微审批详情，计算理论审批人并对比实际审批链。WRULE1 是只读校验器，不新增 HR/考勤审批引擎，不写企微，不替代 WAPPR1 的后台人工确认和真机提交页确认。
 **WTCLOSE2/WTCLOSE2F 执行口径**：新增 `scripts/qa/run-wecom-test-account-final-closeout.sh` 汇总测试号最终证据，复跑 WAUTO1、WFULLSMOKE1、外出/加班 dry-run 和 WRULE1。由于企业微信 API 已配置可信 IP，只允许部署服务器访问，WTCLOSE2 运行证据必须来自部署服务器：本地直接执行脚本默认拒绝跑企微链路，正确命令为 `bash scripts/qa/run-wecom-test-account-final-closeout.sh --remote`，由 SSH 到部署服务器并在 server 容器中执行；本地只做合同、语法、文档和静态门禁。WFULLSMOKE1 在 WTCLOSE2 中以 `WFULLSMOKE_ALLOW_PUBLIC_WRITES=0` 运行，不新增 OA 测试业务数据；如子 runner 触发既有归档写库，只能走 WARCH3/WCHECK1 已批准的 OA 自建留存路径。脚本不写企业微信、不创建外出/加班样件、不启用真实员工。WTCLOSE2 结论只进入 WREMOTE1/WFULLSMOKE2 和 WEXT1-FORMAL 问答，不自动提交审批；此前本地缺 env/DB 造成的 WTCLOSE2 阻断不能作为企微业务阻断证据。
 **WREMOTE1/WFULLSMOKE2/WEXT1-FORMAL 执行口径**：WTCLOSE2F 后下一步先 `SYNC_ENV=0` 部署或确认远端代码同步，再从部署服务器执行 `bash scripts/qa/run-wecom-test-account-final-closeout.sh --remote`。全域 smoke 也必须从部署服务器执行：`WFULLSMOKE_ALLOW_PUBLIC_WRITES=1 bash scripts/qa/run-three-day-full-goal-smoke.sh --remote`，允许创建带 `TEST-SMOKE-WFULL` 标记的 OA 测试业务数据，不写企业微信。外出/加班不再自动创建 OA-POC 测试模板；改为使用现有正式模板，由 `SMOKE002` 真机人工提交，`SMOKE101` 先看详情再审批，系统只读补偿拉取、归档和 WRULE1 校验。WREAL1 真实员工和审批人只通过 `/tmp/hyb-oa-wreal1.env` 临时文件提供，不写入仓库文档或聊天。
+**WEXT1F 执行口径**：正式模板外出/加班证据由 `scripts/qa/run-wecom-formal-outing-overtime-evidence.sh --remote` 从部署服务器执行。该脚本只在用户完成 `SMOKE002` 人工提交、`SMOKE101` 先看详情再处理后运行；远端命令还必须带 `WEXT1F_MANUAL_CONFIRMATION=SMOKE002_SUBMITTED_SMOKE101_REVIEWED`，否则只输出 `PAUSED_MANUAL_CONFIRMATION`，不跑企微 API。它不创建模板、不自动提交审批、不写企业微信，只使用 `--target=outing --target=overtime` 的补偿拉取读取详情，并强制带 `--filter-creator-employee-no=SMOKE002`，避免按正式模板 ID 宽泛读取其他员工单据；在 `WECOM_ARCHIVE_WRITE=1` + `--persist-to-db` 双开关下写 OA 自建留存，再执行外出/加班离线归档读取和 WRULE1。报告结论限定为 `PASS_FORMAL_OUTING_OVERTIME_READY`、`PASS_WITH_FINDINGS`、`PAUSED_MANUAL_CONFIRMATION` 或 `SKIP/BLOCKED_BY_EXTERNAL_CONFIG`；未形成外出/加班正式模板归档前不得进入 WREAL1。
 
 ## 1. 路线图原则
 
@@ -110,7 +111,7 @@
 | 26 | `WREAL1-PREP / WROUTE2 真实试用准备和路线锁定` | 固化候选员工、审批人、允许流程、暂停条件、回滚联系人和验收标准，不启用真实员工，并锁定测试号收口到最小完整镜像的后续顺序。 | 用户选择先收口测试账号，不提前启用真实员工。 | 文档和合同明确真实试用需等四类测试号证据、WAPPR1/WRULE1 审批规则门禁和 WFULLSMOKE1 可解释通过后再问；当前继续测试账号，首批真实试用固定为 1 名真实员工 + 1 名审批人/管理员观察。 |
 | 27 | `WTCLOSE2/WTCLOSE2F 测试号最终收口与证据复核` | 在部署服务器可信 IP 上自动复跑测试号收口证据，确认是否可以进入 WREMOTE1/WFULLSMOKE2 和 WEXT1-FORMAL 问答。 | WAPPR1 四类模板均为 A；WRULE1 已实现；WAUTO1/WEXT1LD/WFULLSMOKE1 已存在；远端代码与本地合同同步。 | 通过 `bash scripts/qa/run-wecom-test-account-final-closeout.sh --remote` 输出 `PASS_TEST_ACCOUNT_CLOSEOUT_READY`、`PASS_WITH_FINDINGS`、`PAUSED_MANUAL_CONFIRMATION` 或 `SKIP/BLOCKED_BY_EXTERNAL_CONFIG`；不写企业微信、不创建样件、不启用真实员工；本地直接执行不作为业务证据。 |
 | 28 | `WREMOTE1/WFULLSMOKE2 远端收口和全域 smoke` | 从部署服务器可信 IP 执行 WTCLOSE2，再远端运行允许 `TEST-SMOKE-WFULL` OA 测试数据的全域 smoke。 | WTCLOSE2F 已完成；允许 `SYNC_ENV=0` 部署；用户批准 OA 主办域公网测试数据。 | 远端报告输出 `PASS_TEST_ACCOUNT_CLOSEOUT_READY`、`PASS_FULL_GOAL_SMOKE_READY`、`PASS_WITH_FINDINGS`、`PAUSED_MANUAL_CONFIRMATION` 或真实 `SKIP/BLOCKED_BY_EXTERNAL_CONFIG`；不写企业微信、不启用真实员工。 |
-| 29 | `WEXT1-FORMAL 正式模板外出/加班测试` | 使用现有正式外出/加班模板，由测试账号人工提交，系统只读校验补偿、归档和审批链。 | WREMOTE1/WFULLSMOKE2 可解释通过；再次问答确认人工提交窗口。 | 不创建 OA-POC 模板、不自动提交审批单；`SMOKE002` 提交，`SMOKE101` 先看详情再审批；若可改审批人、自审或审批链不符，阻断 WREAL1。 |
+| 29 | `WEXT1-FORMAL / WEXT1F 正式模板外出/加班测试` | 使用现有正式外出/加班模板，由测试账号人工提交，系统只读校验补偿、归档和审批链。 | WREMOTE1/WFULLSMOKE2 可解释通过；再次问答确认人工提交窗口；证据脚本必须从部署服务器执行。 | 不创建 OA-POC 模板、不自动提交审批单；`SMOKE002` 提交，`SMOKE101` 先看详情再审批；`run-wecom-formal-outing-overtime-evidence.sh --remote` 输出脱敏报告；若可改审批人、自审或审批链不符，阻断 WREAL1。 |
 | 30 | `WREAL1 小范围真实试用` | 在 1 名真实员工 + 1 名审批人/管理员观察范围验证企微主办 HR/考勤和 OA 留存。 | 请假、补卡、外出、加班四类测试号证据通过；WAPPR1/WRULE1 规则门禁通过；WREMOTE1/WFULLSMOKE2/WEXT1-FORMAL 可解释通过；用户通过 `/tmp/hyb-oa-wreal1.env` 提供真实员工和审批人。 | 试用报告可解释通过或暂停；仍不直接全员推广。 |
 | 31 | `wecom_attendance_approval_mirror_v1 最小完整镜像` | 企微打卡、请假、补卡、外出、加班、假期余额和审批历史镜像到 OA。 | WREAL1 通过后优先规划；不得仅凭测试号证据进入，也不先做应急接管、恢复回写或 RuleOps。 | OA 可查询企微 HR/考勤镜像、附件摘要、审批历史、考勤解释和审计；不把 OA 主办高敏审批迁到企微。 |
 | 32 | `oa_hr_emergency_mode_v1` | 企业微信不可用时，由管理员人工开启 OA HR/考勤应急接管。 | 自建历史留存和镜像数据可支撑对账；OA 应急入口、权限和审计方案明确。 | 管理员可填写原因、范围、预计结束时间和责任人后开启/关闭；应急单据带 `EMERGENCY` 标记。 |
diff --git a/docs/文档事实源状态矩阵.md b/docs/文档事实源状态矩阵.md
index c7f2ddce..d71f225b 100644
--- a/docs/文档事实源状态矩阵.md
+++ b/docs/文档事实源状态矩阵.md
@@ -107,6 +107,7 @@
 | `WTCLOSE2` | `issues/2026-05-22_10-21-12-wecom-test-account-final-closeout-v1.csv` | `CURRENT` | 测试号最终收口与证据复核合同：WAPPR1 四类模板均为 A 后，自动复跑 `/ready`、WAUTO1、WFULLSMOKE1、外出/加班 dry-run 和 WRULE1；不写企业微信、不创建模板或审批单、不启用真实员工；WTCLOSE2F 后只进入 WREMOTE1/WEXT1-FORMAL 问答 |
 | `WTCLOSE2F` | `issues/2026-05-22_11-10-06-wecom-test-account-closeout-remote-only-v1.csv` | `CURRENT` | WTCLOSE2 远端可信 IP 执行口径修正：企微 API 只能由部署服务器访问，WTCLOSE2 真实运行必须使用 `--remote` 在 server 容器内执行；本地直接执行默认阻断，只保留静态门禁用途 |
 | `WREMOTE1` | `issues/2026-05-22_12-34-41-wecom-remote-closeout-and-full-smoke-v1.csv` | `CURRENT` | 远端可信 IP 收口、WFULLSMOKE2 全域 smoke 和 WEXT1-FORMAL 正式模板外出/加班测试合同：先远端 WTCLOSE2，再远端全域 smoke；外出/加班不自动建模板，改为 `SMOKE002` 正式模板人工提交、`SMOKE101` 审批后系统只读补偿/归档/WRULE1 |
+| `WEXT1F` | `issues/2026-05-22_14-23-19-wecom-formal-outing-overtime-evidence-v1.csv` | `CURRENT` | 正式模板外出/加班证据脚本合同：`SMOKE002` 人工提交、`SMOKE101` 先看详情再审批后，使用 `WEXT1F_MANUAL_CONFIRMATION=SMOKE002_SUBMITTED_SMOKE101_REVIEWED bash scripts/qa/run-wecom-formal-outing-overtime-evidence.sh --remote` 从部署服务器只读拉取 outing/overtime；补偿拉取必须按 `creator=SMOKE002` 过滤，只写 OA 留存、离线读取和 WRULE1；不创建模板、不自动提交审批、不写企业微信 |
 | `WREAL1-PREP` | `issues/2026-05-20_14-35-42-wecom-real-trial-prep-v1.csv` | `CURRENT` | 真实小范围试用准备合同：只固化候选员工、非自审审批人、回滚联系人、允许流程、暂停条件和验收标准；当前不启用真实员工，真实试用必须等请假/补卡/外出/加班四类测试号证据、WAPPR1 审批规则门禁和 WFULLSMOKE1 全域主路径可解释通过后再问 |
 
 旧 Issue CSV 保留为历史审计证据：不删除、不批量回写、不移动到 `docs/archive/` 或其他归档目录。新连续合同如替代旧合同，只在新合同 `Notes` 或文档索引中说明 supersedes。
diff --git a/docs/文档索引.md b/docs/文档索引.md
index e65f5544..c3edb8b3 100644
--- a/docs/文档索引.md
+++ b/docs/文档索引.md
@@ -20,7 +20,7 @@
 
 ## 最近事实源链
 
-R6 -> PC6 -> PA6 -> FL6 -> AP6 -> RI6 -> LP6 -> ME6/QG6 -> DG7/MRF7 -> MIA7 -> MIF7/MIF7T(公网部署与带凭据公网复检已收口) -> DG8 -> DG9 -> R7E -> FBA -> QAOPS9/P2D -> P2-ORG -> P2-FIN-DECISION -> P2-FIN-V1 -> P2-OCR-SAMPLE -> P2Q -> P2AUTO -> P2AF -> P2-FIN-LIVE-DECISION -> P2-FIN-LIVE -> PRELAUNCH -> CASHIER-VOUCHER-V1 -> LAUNCH-DEPLOY-FINANCE -> FIN-PROD-ENABLE -> ACCOUNTING-HANDOFF-V1 -> SYSTEM-COMPLETE-GATE -> LAUNCH-OPS-V1 -> REQ-COMPLETE-AUDIT -> ATTV1 -> LAUNCH-AUTO-OPS-V1 -> AUTO-REMAINING-CLOSEOUT -> HR-WECOM-DIRECTORY-V1 -> HRWD-CONTACT-LIVE-TEST -> WCFB -> WFOC -> DBWFOC -> WFDG -> WDRG -> WCP1 -> WCPC -> WHMS -> WHMSLIVE -> WOUA -> WAPI -> WRRA -> WOUA-FIX -> WRV1 -> WRA1 -> WMSG1 -> WAPC1 -> WAPC2 -> WPROC1 -> WPROC2 -> WAPC2F dry-run -> WAPC2FLIVE -> WCBK1 -> WEGOV1 -> WACP3/WACP3R -> WREAD1 自动汇总 -> WREAD1F 新样例 -> WREAD1 真机确认 -> WADM1 自建留存模型设计 -> WARCH1 最小审批历史留存 -> WARCH2 正式只读历史查询 -> WROAD2+OPS 受控上线与每日巡检 -> WARCH3 归档稳定化自动巡检 -> WCHECK1 打卡摘要留存 -> WEXT1 外出/加班 dry-run -> WOPS1 巡检告警降噪 -> WROAD2-CLOSEOUT 三天验收 -> WROAD2-BIZ1 三天业务冒烟 -> WAPPR1 审批规则规范化 -> WRULE1 审批链校验器 POC -> WAUTO1 自动证据汇总 -> WEXT1LD live 前 dry-run -> WFULLSMOKE1 三天全域最终目标测试 -> WROUTE2 测试号收口到最小镜像路线锁定 -> WTCLOSE1 小团队自动审批规则与测试号全域收口 dry-run -> WTCLOSE2 测试号最终收口与证据复核 -> WTCLOSE2F 远端可信 IP 执行口径修正 -> WREMOTE1 远端可信 IP 收口 -> WFULLSMOKE2 全域 smoke -> WEXT1-FORMAL 正式模板外出/加班测试 -> 四类测试号证据和审批规则门禁通过后问答 WREAL1 小范围真实试用（首批 1 名真实员工 + 1 名审批人/管理员观察） -> wecom_attendance_approval_mirror_v1 最小完整镜像。
+R6 -> PC6 -> PA6 -> FL6 -> AP6 -> RI6 -> LP6 -> ME6/QG6 -> DG7/MRF7 -> MIA7 -> MIF7/MIF7T(公网部署与带凭据公网复检已收口) -> DG8 -> DG9 -> R7E -> FBA -> QAOPS9/P2D -> P2-ORG -> P2-FIN-DECISION -> P2-FIN-V1 -> P2-OCR-SAMPLE -> P2Q -> P2AUTO -> P2AF -> P2-FIN-LIVE-DECISION -> P2-FIN-LIVE -> PRELAUNCH -> CASHIER-VOUCHER-V1 -> LAUNCH-DEPLOY-FINANCE -> FIN-PROD-ENABLE -> ACCOUNTING-HANDOFF-V1 -> SYSTEM-COMPLETE-GATE -> LAUNCH-OPS-V1 -> REQ-COMPLETE-AUDIT -> ATTV1 -> LAUNCH-AUTO-OPS-V1 -> AUTO-REMAINING-CLOSEOUT -> HR-WECOM-DIRECTORY-V1 -> HRWD-CONTACT-LIVE-TEST -> WCFB -> WFOC -> DBWFOC -> WFDG -> WDRG -> WCP1 -> WCPC -> WHMS -> WHMSLIVE -> WOUA -> WAPI -> WRRA -> WOUA-FIX -> WRV1 -> WRA1 -> WMSG1 -> WAPC1 -> WAPC2 -> WPROC1 -> WPROC2 -> WAPC2F dry-run -> WAPC2FLIVE -> WCBK1 -> WEGOV1 -> WACP3/WACP3R -> WREAD1 自动汇总 -> WREAD1F 新样例 -> WREAD1 真机确认 -> WADM1 自建留存模型设计 -> WARCH1 最小审批历史留存 -> WARCH2 正式只读历史查询 -> WROAD2+OPS 受控上线与每日巡检 -> WARCH3 归档稳定化自动巡检 -> WCHECK1 打卡摘要留存 -> WEXT1 外出/加班 dry-run -> WOPS1 巡检告警降噪 -> WROAD2-CLOSEOUT 三天验收 -> WROAD2-BIZ1 三天业务冒烟 -> WAPPR1 审批规则规范化 -> WRULE1 审批链校验器 POC -> WAUTO1 自动证据汇总 -> WEXT1LD live 前 dry-run -> WFULLSMOKE1 三天全域最终目标测试 -> WROUTE2 测试号收口到最小镜像路线锁定 -> WTCLOSE1 小团队自动审批规则与测试号全域收口 dry-run -> WTCLOSE2 测试号最终收口与证据复核 -> WTCLOSE2F 远端可信 IP 执行口径修正 -> WREMOTE1 远端可信 IP 收口 -> WFULLSMOKE2 全域 smoke -> WEXT1-FORMAL/WEXT1F 正式模板外出/加班测试 -> 四类测试号证据和审批规则门禁通过后问答 WREAL1 小范围真实试用（首批 1 名真实员工 + 1 名审批人/管理员观察） -> wecom_attendance_approval_mirror_v1 最小完整镜像。
 
 - R6、PC6、PA6、FL6、AP6、RI6、LP6 已完成并部署。
 - ME6 自动化、公网视觉矩阵和 `ME6-3` 代表性企业微信真机门禁已完成。
@@ -168,6 +168,7 @@ R6 -> PC6 -> PA6 -> FL6 -> AP6 -> RI6 -> LP6 -> ME6/QG6 -> DG7/MRF7 -> MIA7 -> M
 | `WTCLOSE2` | `issues/2026-05-22_10-21-12-wecom-test-account-final-closeout-v1.csv` | 当前测试号最终收口与证据复核：WEXT1-FORMAL 前最后自动收口 | 新增 `scripts/qa/run-wecom-test-account-final-closeout.sh`，复跑 `/ready`、WAUTO1、WFULLSMOKE1、外出/加班 dry-run 和 WRULE1；WFULLSMOKE1 在该脚本中关闭公网写流程，WRULE1 使用 `--manual-gate=all:A` | 不写企业微信；不新增 OA 测试业务数据；不创建外出/加班模板或审批单；不启用真实员工；结论只进入 WEXT1-FORMAL 问答 |
 | `WTCLOSE2F` | `issues/2026-05-22_11-10-06-wecom-test-account-closeout-remote-only-v1.csv` | 当前 WTCLOSE2 远端可信 IP 执行口径修正 | WTCLOSE2 真实运行命令改为 `bash scripts/qa/run-wecom-test-account-final-closeout.sh --remote`，由部署服务器 server 容器执行；本地直接执行默认拒绝跑企微链路 | 本地只做合同、语法、文档和静态门禁；此前本地缺 env/DB 的 WTCLOSE2 输出不作为企微业务阻断证据 |
 | `WREMOTE1` | `issues/2026-05-22_12-34-41-wecom-remote-closeout-and-full-smoke-v1.csv` | 当前远端收口、全域 smoke 和正式模板外出/加班测试路线 | WFULLSMOKE2 必须使用 `--remote` 在部署服务器执行；外出/加班使用正式模板由 `SMOKE002` 人工提交、`SMOKE101` 审批，系统只读补偿、归档和 WRULE1 | 允许 OA `TEST-SMOKE-WFULL` 测试数据；不写企业微信、不自动创建外出/加班模板或审批单；真实员工信息只通过 `/tmp/hyb-oa-wreal1.env` 临时文件 |
+| `WEXT1F` | `issues/2026-05-22_14-23-19-wecom-formal-outing-overtime-evidence-v1.csv` | 当前正式模板外出/加班证据脚本：人工提交后从部署服务器只读收证据 | 新增 `WEXT1F_MANUAL_CONFIRMATION=SMOKE002_SUBMITTED_SMOKE101_REVIEWED bash scripts/qa/run-wecom-formal-outing-overtime-evidence.sh --remote`；显式拉取 `outing/overtime` 且按 `creator=SMOKE002` 过滤，写 OA 留存仅限 `WECOM_ARCHIVE_WRITE=1` + `--persist-to-db`；随后做离线归档读取和 WRULE1 | 不创建模板、不自动提交审批、不写企业微信、不启用真实员工；未有人工作业或缺确认标记时只可得到人工缺口结论 |
 | `WREAL1-PREP` | `issues/2026-05-20_14-35-42-wecom-real-trial-prep-v1.csv` | 当前真实试用准备：不启用真实员工 | 固化候选员工、非自审审批人、回滚联系人、允许流程、暂停条件和验收标准；真实试用必须等请假、补卡、外出、加班四类测试号证据、WAPPR1 和 WFULLSMOKE1 通过后再问 | 当前不创建外出/加班 live 样件；真实员工名单不写入 Issue 或文档，后续通过问答确认 |
 
 旧 Issue CSV 不删除、不批量回写、不移动到归档目录。若新治理流替代旧合同，只在新 CSV 的 `Notes` 或本索引说明 supersedes，旧文件保留历史审计语义。
diff --git a/docs/管理员手册.md b/docs/管理员手册.md
index 6a8709c7..f34a8fe0 100644
--- a/docs/管理员手册.md
+++ b/docs/管理员手册.md
@@ -7,6 +7,7 @@
 **WAPPR1 2026-05-21 真机结论**：测试员工直属上级已设置为测试主管；请假、补卡、外出、加班提交页均自动带出测试主管，且申请人不能删除、改人或新增审批人，四类模板均按 A 记录。该结论仍需结合提交后的 WRULE1 实际审批链校验和三天业务冒烟证据，不单独放行真实员工试用。
 **WRULE1 审批链校验**：管理员后台规则和真机提交页仍由 WAPPR1 人工确认；系统只能用已归档审批详情校验实际审批链。校验命令为 `npm run check:wecom-approval-rule-validator -- --template=<类型> --manual-gate=<类型>:A`，缺人工门禁时结论为 `PAUSED_MANUAL_CONFIRMATION`；自审、实际审批人不符或申请人可手选审批人均阻断 WEXT1-FORMAL、WREAL1 和最小镜像。
 **WTCLOSE2/WTCLOSE2F 测试号最终收口**：企业微信 API 已限制只能由部署服务器可信 IP 访问，因此命令必须使用 `bash scripts/qa/run-wecom-test-account-final-closeout.sh --remote`。本地直接执行脚本会拒绝跑企微链路，只能用于 `bash -n`、合同和文档静态门禁；不得把本地缺 env/DB 的结果当作业务阻断。远端脚本复跑 WAUTO1、WFULLSMOKE1、外出/加班 dry-run 和 WRULE1，并汇总 WROAD2/WARCH3 既有报告；WFULLSMOKE1 在该脚本中关闭公网业务写流程，不新增 OA 测试业务数据；如子 runner 触发既有归档写库，只能走 WARCH3/WCHECK1 已批准的 OA 自建留存路径；脚本不写企业微信、不创建审批样件、不启用真实员工。
+**WEXT1F 正式模板外出/加班证据脚本**：管理员或 QA 只在人工测试完成后执行 `WEXT1F_MANUAL_CONFIRMATION=SMOKE002_SUBMITTED_SMOKE101_REVIEWED bash scripts/qa/run-wecom-formal-outing-overtime-evidence.sh --remote`。人工步骤固定为：`SMOKE002` 在企业微信正式外出模板提交低风险测试单，确认审批人自动带出测试主管且不能改人；再在正式加班模板提交低风险测试单并做同样确认；`SMOKE101` 必须先打开详情页再同意或驳回，不能直接点快捷同意。缺少该人工确认标记时脚本只输出 `PAUSED_MANUAL_CONFIRMATION`，不跑企微 API。脚本从部署服务器可信 IP 读取企微审批 API，不创建模板、不自动提交审批、不写企业微信；补偿拉取必须同时按 `template_id` 和 `creator=SMOKE002` 过滤，避免读取同一正式模板下其他员工单据；仅在既有 `WECOM_ARCHIVE_WRITE=1` + `--persist-to-db` 双开关下把拉到的外出/加班详情写入 OA 自建留存，然后做离线归档读取和 WRULE1 审批链校验。
 
 ## 1. 适用范围
 - 适用于 `small_team_internal` 工作区模式。
diff --git a/docs/需求-统一审批中心规约.md b/docs/需求-统一审批中心规约.md
index afb4282d..aec27c91 100644
--- a/docs/需求-统一审批中心规约.md
+++ b/docs/需求-统一审批中心规约.md
@@ -456,7 +456,11 @@ WTCLOSE2F 后，企业微信审批相关业务证据必须来自部署服务器
 
 - WREMOTE1 先部署/同步代码并执行 `bash scripts/qa/run-wecom-test-account-final-closeout.sh --remote`；全域 smoke 使用 `WFULLSMOKE_ALLOW_PUBLIC_WRITES=1 bash scripts/qa/run-three-day-full-goal-smoke.sh --remote`。
 - WEXT1-FORMAL 人工操作为 `SMOKE002` 在正式外出、加班模板提交低风险测试单，`SMOKE101` 先打开详情页再审批。
+- 系统只读校验 runner 为 `WEXT1F_MANUAL_CONFIRMATION=SMOKE002_SUBMITTED_SMOKE101_REVIEWED bash scripts/qa/run-wecom-formal-outing-overtime-evidence.sh --remote`；只能从部署服务器可信 IP 执行。本地直接执行必须拒绝业务运行，缺少人工确认标记时不得跑企微 API。
+- 该 runner 显式使用 `--target=outing --target=overtime` 拉取正式模板外出/加班详情，并同时设置 `--filter-creator-employee-no=SMOKE002`；默认请假/补卡巡检范围不得因此扩大，也不得按正式模板 ID 宽泛读取同模板其他员工单据。
+- OA 归档写入只允许走既有双开关：`WECOM_ARCHIVE_WRITE=1` + `--persist-to-db`；这只写 OA 自建留存库，不写企业微信。
 - 系统只读校验为补偿拉取、审批详情读取、OA 归档读取、WRULE1 审批链校验和隐私扫描；不得调用创建模板、提交审批或其他企微写接口。
+- WEXT1F 结论限定为 `PASS_FORMAL_OUTING_OVERTIME_READY`、`PASS_WITH_FINDINGS`、`PAUSED_MANUAL_CONFIRMATION` 或 `SKIP/BLOCKED_BY_EXTERNAL_CONFIG`。
 - 若正式模板出现申请人可任意改审批人、自审、审批人不一致、可见范围或可信 IP 阻断，结论记录为 blocking finding，不进入 WREAL1。
 - WREAL1 真实员工和审批人信息只通过 `/tmp/hyb-oa-wreal1.env` 临时文件提供，不写入 Issue CSV、仓库文档或聊天。
 
diff --git a/issues/2026-05-22_14-23-19-wecom-formal-outing-overtime-evidence-v1.csv b/issues/2026-05-22_14-23-19-wecom-formal-outing-overtime-evidence-v1.csv
new file mode 100644
index 00000000..a49cb47c
--- /dev/null
+++ b/issues/2026-05-22_14-23-19-wecom-formal-outing-overtime-evidence-v1.csv
@@ -0,0 +1,9 @@
+"ID","Title","Description","Acceptance","Test_Method","Tools","Dev_Status","Review1_Status","Regression_Status","Files","Dependencies","Notes"
+"WEXT1F-0","建立正式模板外出加班证据合同","建立 WEXT1-FORMAL 执行合同，承接 WREMOTE1 后的路线：不再自动创建 OA-POC 外出/加班模板，不自动提交审批单；由 SMOKE002 在企业微信正式模板人工提交外出和加班，SMOKE101 先看详情再审批，系统只从部署服务器可信 IP 做只读拉取、OA 归档和审批链校验；审批读取 Secret 来源固定 WECOM_AGENT_SECRET，不尝试其他 Secret、无 fallback、不允许运行时猜 Secret。","1) Issue ID 使用 WEXT1F-0..WEXT1F-7；2) 明确企业微信 API 证据只从部署服务器可信 IP 获取；3) 明确本轮不写企业微信、不创建模板、不自动提交审批、不启用真实员工；4) 明确 OA 写入只限 WECOM_ARCHIVE_WRITE=1 + --persist-to-db 的既有归档路径。","受影响流程：正式模板外出/加班人工提交 -> 企微补偿只读拉取 -> OA 归档 -> WRULE1 审批链校验 -> WREAL1 准入；验证步骤：python3 /home/shunli/.codex/skills/plan/scripts/validate_issues_csv.py issues/2026-05-22_14-23-19-wecom-formal-outing-overtime-evidence-v1.csv；node scripts/qa/check-process-contract.mjs --issue issues/2026-05-22_14-23-19-wecom-formal-outing-overtime-evidence-v1.csv --profile wecom --json。","python3|node","DONE","DONE","DONE","issues/2026-05-22_14-23-19-wecom-formal-outing-overtime-evidence-v1.csv","[]","回滚/清理：删除本合同；不修改 Prisma schema；不写企业微信。"
+"WEXT1F-1","扩展补偿拉取显式目标","扩展 check-wecom-approval-compensation-poc，使其默认仍只覆盖请假和补卡，但支持显式 --target=outing --target=overtime；支持从正式模板提交记录按 SMOKE002 创建人只读发现外出/加班模板 ID，并在正式模板详情拉取时强制使用 --filter-creator-employee-no=SMOKE002，避免按 template_id 宽泛读取同模板其他员工单据。","1) 默认无 --target 时仍为 leave + missed_punch；2) 显式 --target=outing --target=overtime 才纳入外出/加班；3) 支持 WECOM_APPROVAL_OUTING_TEMPLATE_ID/WECOM_APPROVAL_OVERTIME_TEMPLATE_ID 或 --discover-template-ids 发现；4) WEXT1F 正式模板拉取必须带 creator=SMOKE002 过滤，无法解析则不拉取；5) 报告只输出脱敏前缀和计数；6) 不新增企微写入开关。","受影响流程：正式模板人工提交后 -> getapprovalinfo 按创建人发现模板 -> 按 template_id + creator 拉取外出/加班详情；验证步骤：cd server && npm run test -- tests/unit/scripts/check-wecom-approval-compensation-poc.test.js。","node|vitest","DONE","DONE","DONE","server/src/scripts/check-wecom-approval-compensation-poc.js|server/src/services/integrations/wecomApprovalArchive.js|server/tests/unit/scripts/check-wecom-approval-compensation-poc.test.js","[""WEXT1F-0""]","回滚/清理：恢复补偿 runner；默认范围保持不变，避免误扩大 WARCH3/WAUTO1 现有巡检。"
+"WEXT1F-2","扩展 OA 归档离线读取目标","扩展 check-wecom-approval-archive 离线读取目标，支持 --template=outing 和 --template=overtime，用于证明正式模板外出/加班详情已经进入 OA 自建留存。","1) 默认离线读取仍为 leave + missed_punch；2) 显式模板参数可读 outing/overtime；3) 不调用企微 API；4) 报告不输出完整 sp_no、template_id、userid、表单值、附件、评论或原始 payload。","受影响流程：OA 归档库 -> 外出/加班历史只读查看；验证步骤：cd server && npm run test -- tests/unit/scripts/check-wecom-approval-archive.test.js。","node|vitest","DONE","DONE","DONE","server/src/scripts/check-wecom-approval-archive.js|server/tests/unit/scripts/check-wecom-approval-archive.test.js","[""WEXT1F-1""]","回滚/清理：恢复离线读取目标集合；不改业务 API 或前端页面。"
+"WEXT1F-3","新增远端正式模板证据脚本","新增 QA-only 脚本 scripts/qa/run-wecom-formal-outing-overtime-evidence.sh；本地默认拒绝业务执行，--remote 通过 SSH 到部署服务器容器运行；容器内只读拉取外出/加班、可选归档到 OA、自建留存离线读、WRULE1 校验和隐私扫描；脚本必须要求人工确认标记 WEXT1F_MANUAL_CONFIRMATION=SMOKE002_SUBMITTED_SMOKE101_REVIEWED，否则只输出 PAUSED，不跑企微 API。","1) 本地直接执行必须提示 --remote；2) 即使伪造远端环境变量，本地非 /app 容器路径也必须拒绝；3) 远端执行使用 WEXT1F_EXECUTION_MODE=remote 和 WEXT1F_REMOTE_CONTEXT=deployment-server；4) 缺人工确认标记时不跑企微 API；5) 不设置任何企微写入开关；6) 归档写 OA 只通过 WECOM_ARCHIVE_WRITE=1 + --persist-to-db；7) 输出 artifacts/qa-flow/wecom-formal-outing-overtime-evidence-<timestamp>/summary.json。","受影响流程：正式模板人工提交后 -> 远端只读收证据；验证步骤：bash -n scripts/qa/run-wecom-formal-outing-overtime-evidence.sh；bash scripts/qa/run-wecom-formal-outing-overtime-evidence.sh --help；本地直接执行返回非 0 且提示 --remote；WEXT1F_EXECUTION_MODE=remote WEXT1F_REMOTE_CONTEXT=deployment-server bash scripts/qa/run-wecom-formal-outing-overtime-evidence.sh 仍应因非 /app 容器路径拒绝。","bash|ssh|node","DONE","DONE","DONE","scripts/qa/run-wecom-formal-outing-overtime-evidence.sh","[""WEXT1F-2""]","回滚/清理：删除脚本；脚本不创建模板、不提交审批、不启用真实员工。"
+"WEXT1F-4","更新事实源和人工步骤","更新现有事实源，记录 WEXT1-FORMAL 的正式模板人工操作清单、可信 IP 边界、外出/加班证据准入和 WREAL1 后置条件；不新建独立说明文档。","1) 文档明确 SMOKE002 人工提交正式模板外出/加班；2) 文档明确 SMOKE101 先看详情再审批；3) 文档明确系统只能只读拉取、归档和校验；4) 文档明确 WREAL1 仍需另行问答。","受影响流程：管理员按文档执行 WEXT1-FORMAL；验证步骤：rg -n ""WEXT1-FORMAL|正式模板|SMOKE002|SMOKE101|run-wecom-formal-outing-overtime-evidence|WREAL1"" docs .agents/skills/hyb-oa-delivery/SKILL.md；node scripts/qa/audit-docs.mjs --check-registry；quick_validate.py .agents/skills/hyb-oa-delivery。","rg|node|python3","DONE","DONE","DONE","docs/企微优先开发路线图.md|docs/管理员手册.md|docs/trialrun/人工试运行报告.md|docs/文档索引.md|docs/文档事实源状态矩阵.md|docs/需求-统一审批中心规约.md|.agents/skills/hyb-oa-delivery/SKILL.md","[""WEXT1F-3""]","回滚/清理：恢复文档和 skill；不记录真实员工信息。"
+"WEXT1F-5","本地门禁与只读评审","执行合同校验、process gate、脚本语法、focused 单测、文档审计、skill 校验、diff 检查和最小回归；涉及审批规则和权限边界，实施后启动只读 reviewer subagent。","1) validate_issues_csv 通过；2) baseline 和 issue-specific process gate 通过；3) bash -n 通过；4) focused vitest 通过；5) docs/skill/diff/minimal gates 通过；6) reviewer 无 blocker 或 blocker 已修复。","受影响流程：脚本/规则边界 -> 独立评审 -> 部署前门禁；验证步骤：validate_issues_csv.py <csv>；node scripts/qa/check-process-contract.mjs --baseline --json；node scripts/qa/check-process-contract.mjs --issue <csv> --profile wecom --json；cd server && npm run test -- tests/unit/scripts/check-wecom-approval-compensation-poc.test.js tests/unit/scripts/check-wecom-approval-archive.test.js tests/unit/scripts/check-wecom-approval-rule-validator.test.js；bash scripts/check-minimal.sh。","python3|node|bash|subagent|git","DONE","DONE","DONE","server/src/scripts/*|scripts/qa/run-wecom-formal-outing-overtime-evidence.sh|docs/*|.agents/skills/hyb-oa-delivery/SKILL.md","[""WEXT1F-4""]","本地门禁已通过；只读 reviewer 已复核 creator 过滤、本地伪远端拒绝和人工确认门禁无 blocker；部署/人工提交/远端取证仍在 WEXT1F-6/7。"
+"WEXT1F-6","部署并等待人工提交","本地门禁通过后 SYNC_ENV=0 部署，复检 /ready；然后给出 SMOKE002/SMOKE101 真机操作步骤，等待用户确认已提交/已审批后再跑远端证据脚本。","1) SYNC_ENV=0 部署成功；2) /ready 正常；3) 不在用户人工提交前跑正式模板证据通过结论；4) 给出按账号、模板、时间、确认点排列的人工步骤。","受影响流程：部署 -> 人工正式模板外出/加班 -> 远端证据脚本；验证步骤：SYNC_ENV=0 bash deploy/trialrun/deploy.sh；curl -fsS https://oa.hyinda.com/ready；人工回复已提交/已审批。","bash|curl|manual","TODO","TODO","TODO","deploy/trialrun/*|scripts/qa/run-wecom-formal-outing-overtime-evidence.sh","[""WEXT1F-5""]","回滚/清理：若用户暂停，不跑远端证据；已部署代码可由后续提交回滚。"
+"WEXT1F-7","远端收证据并回填提交","用户完成正式模板外出/加班人工测试后，从部署服务器执行 WEXT1F 证据脚本，回填报告路径和真实结论，中文提交。","1) 远端 summary 结论限定 PASS_FORMAL_OUTING_OVERTIME_READY、PASS_WITH_FINDINGS、PAUSED_MANUAL_CONFIRMATION、SKIP/BLOCKED_BY_EXTERNAL_CONFIG；2) 报告包含补偿拉取、OA 归档离线读、WRULE1、隐私扫描；3) 不泄露完整 sp_no/template_id/userid/手机号/邮箱/token/Secret/EncodingAESKey/表单值/附件/评论/原始 payload；4) commit message 使用中文并引用 WEXT1F。","受影响流程：正式模板外出/加班证据 -> WREAL1 问答准入；验证步骤：bash scripts/qa/run-wecom-formal-outing-overtime-evidence.sh --remote；git status --short；git log -1 --oneline。","bash|ssh|git","TODO","TODO","TODO","issues/2026-05-22_14-23-19-wecom-formal-outing-overtime-evidence-v1.csv|server/src/scripts/*|scripts/qa/run-wecom-formal-outing-overtime-evidence.sh|docs/*|.agents/skills/hyb-oa-delivery/SKILL.md","[""WEXT1F-6""]","回滚/清理：不删除远端审计报告；若正式模板不满足 A 或实际审批链不一致，记录 blocker 并不进入 WREAL1。"
diff --git a/scripts/qa/run-wecom-formal-outing-overtime-evidence.sh b/scripts/qa/run-wecom-formal-outing-overtime-evidence.sh
new file mode 100755
index 00000000..d3fa2dc3
--- /dev/null
+++ b/scripts/qa/run-wecom-formal-outing-overtime-evidence.sh
@@ -0,0 +1,341 @@
+#!/usr/bin/env bash
+set -u -o pipefail
+
+ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/../.." && pwd)"
+SERVER_DIR="$ROOT_DIR/server"
+PUBLIC_BASE_URL="${LIVE_SMOKE_BASE_URL:-https://oa.hyinda.com}"
+REPORT_DIR="${WEXT1F_REPORT_DIR:-$ROOT_DIR/artifacts/qa-flow/wecom-formal-outing-overtime-evidence-$(date +%Y%m%d-%H%M%S)}"
+SUMMARY_TXT="$REPORT_DIR/summary.txt"
+STATUS=0
+REQUIRED_MANUAL_CONFIRMATION="SMOKE002_SUBMITTED_SMOKE101_REVIEWED"
+
+usage() {
+  cat <<'EOF'
+Usage:
+  bash scripts/qa/run-wecom-formal-outing-overtime-evidence.sh --remote
+
+WEXT1-FORMAL must run on the deployment server because Enterprise WeChat API
+access is restricted by trusted IP. Local execution is intentionally blocked;
+use bash -n for syntax/static checks.
+
+This runner assumes SMOKE002 has manually submitted formal outing/overtime
+approval records in Enterprise WeChat. It does not create templates, submit
+approvals, or write Enterprise WeChat. It only reads WeCom approval APIs from
+the deployment server and persists pulled details to the OA archive behind the
+existing WECOM_ARCHIVE_WRITE=1 + --persist-to-db gate.
+EOF
+}
+
+run_remote() {
+  local deploy_server="${DEPLOY_SERVER:-deploy@118.89.77.218}"
+  local remote_dir="${REMOTE_DIR:-/opt/hyb-oa-trialrun}"
+  local remote_env_file="${REMOTE_ENV_FILE:-${remote_dir}/deploy/trialrun/.env}"
+  local compose_file="${COMPOSE_FILE:-deploy/trialrun/docker-compose.yml}"
+  local public_base_url="${LIVE_SMOKE_BASE_URL:-$PUBLIC_BASE_URL}"
+  local manual_confirmation=""
+  if [[ "${WEXT1F_MANUAL_CONFIRMATION:-}" == "$REQUIRED_MANUAL_CONFIRMATION" ]]; then
+    manual_confirmation="$REQUIRED_MANUAL_CONFIRMATION"
+  fi
+
+  ssh "$deploy_server" \
+    "REMOTE_DIR='${remote_dir}' REMOTE_ENV_FILE='${remote_env_file}' COMPOSE_FILE='${compose_file}' LIVE_SMOKE_BASE_URL='${public_base_url}' WEXT1F_MANUAL_CONFIRMATION='${manual_confirmation}' bash -s" <<'REMOTE'
+set -euo pipefail
+cd "$REMOTE_DIR"
+if docker compose version >/dev/null 2>&1; then
+  compose_cmd=(docker compose)
+else
+  compose_cmd=(docker-compose)
+fi
+"${compose_cmd[@]}" -f "$COMPOSE_FILE" --env-file "$REMOTE_ENV_FILE" exec -T server sh -lc \
+  "cd /app && LIVE_SMOKE_BASE_URL='$LIVE_SMOKE_BASE_URL' WEXT1F_MANUAL_CONFIRMATION='$WEXT1F_MANUAL_CONFIRMATION' WEXT1F_EXECUTION_MODE=remote WEXT1F_REMOTE_CONTEXT=deployment-server bash scripts/qa/run-wecom-formal-outing-overtime-evidence.sh"
+REMOTE
+}
+
+case "${1:-}" in
+  --remote)
+    run_remote
+    exit $?
+    ;;
+  --help|-h)
+    usage
+    exit 0
+    ;;
+  "")
+    ;;
+  *)
+    printf '[wext1f][ERROR] unknown argument: %s\n' "$1" >&2
+    usage >&2
+    exit 2
+    ;;
+esac
+
+if [[ "${WEXT1F_EXECUTION_MODE:-}" != "remote" || "${WEXT1F_REMOTE_CONTEXT:-}" != "deployment-server" ]]; then
+  printf '[wext1f][ERROR] Enterprise WeChat API checks must run on the deployment server trusted IP. Use --remote.\n' >&2
+  usage >&2
+  exit 2
+fi
+
+if [[ "$ROOT_DIR" != "/app" ]]; then
+  printf '[wext1f][ERROR] remote execution must run inside the deployment server container at /app. Use --remote.\n' >&2
+  exit 2
+fi
+
+mkdir -p "$REPORT_DIR"
+: >"$SUMMARY_TXT"
+
+log() {
+  printf '%s\n' "$*" | tee -a "$SUMMARY_TXT"
+}
+
+run_step() {
+  local name="$1"
+  shift
+  local log_file="$REPORT_DIR/${name}.log"
+
+  log ">>> [wext1f] $name"
+  if "$@" > >(tee "$log_file") 2>&1; then
+    log "[wext1f][PASS] $name"
+  else
+    local code="$?"
+    log "[wext1f][FAIL] $name exit=$code"
+    STATUS=1
+  fi
+}
+
+run_server_step() {
+  local name="$1"
+  shift
+  local log_file="$REPORT_DIR/${name}.log"
+
+  log ">>> [wext1f] $name"
+  if (cd "$SERVER_DIR" && "$@") > >(tee "$log_file") 2>&1; then
+    log "[wext1f][PASS] $name"
+  else
+    local code="$?"
+    log "[wext1f][FAIL] $name exit=$code"
+    STATUS=1
+  fi
+}
+
+run_ready() {
+  run_step "public-ready" curl -fsS "$PUBLIC_BASE_URL/ready"
+}
+
+run_formal_compensation_archive() {
+  run_server_step "formal-outing-overtime-compensation-archive" env \
+    WECOM_ARCHIVE_WRITE=1 \
+    npm run check:wecom-approval-compensation-poc -- \
+    --live \
+    --format=json \
+    --target=outing \
+    --target=overtime \
+    --discover-template-ids \
+    --discover-applicant-employee-no=SMOKE002 \
+    --filter-creator-employee-no=SMOKE002 \
+    --persist-to-db \
+    --report-dir="$REPORT_DIR/wecom-approval-compensation-poc"
+}
+
+run_archive_offline() {
+  run_server_step "formal-outing-overtime-archive-offline" npm run check:wecom-approval-archive -- \
+    --offline \
+    --format=json \
+    --template=outing \
+    --template=overtime \
+    --limit=50 \
+    --report-dir="$REPORT_DIR/wecom-approval-archive"
+}
+
+run_rule_validator() {
+  run_server_step "formal-outing-overtime-rule-validator" npm run check:wecom-approval-rule-validator -- \
+    --format=json \
+    --template=outing \
+    --template=overtime \
+    --manual-gate=all:A \
+    --limit=50 \
+    --report-dir="$REPORT_DIR/wecom-approval-rule-validator"
+}
+
+run_privacy_smoke() {
+  local sensitive_pattern='(["'\'']?(WECOM_[A-Z0-9_]*(SECRET|TOKEN)|[A-Z0-9_]*(TOKEN|SECRET|PASSWORD|ENCODING_AES_KEY|ENCODINGAESKEY|CORP_SECRET|ACCESS_TOKEN)[A-Z0-9_]*)["'\'']?[[:space:]]*[:=][[:space:]]*["'\'']?[A-Za-z0-9._:/+=-]{8,}["'\'']?)|Authorization:[[:space:]]*Bearer[[:space:]]+[A-Za-z0-9._-]{8,}|access_token=[A-Za-z0-9._-]{8,}|1[3-9][0-9]{9}|[[:alnum:]._%+-]+@[[:alnum:].-]+\.[[:alpha:]]{2,}|-----BEGIN|BEGIN [A-Z ]*KEY|"[^"]*(token|secret|password|encodingAESKey|encoding_aes_key|sp_no|spNo|template_id|templateId|userid|userId|external_userid|externalUserId|external_user_id)[^"]*"[[:space:]]*:[[:space:]]*"[^"]{8,}"|\b20[0-9]{12,}\b|"(raw|rawPayload|raw_payload|payload|xml|ciphertext|Encrypt|formValue|formValues|attachments|comments)"[[:space:]]*:'
+  local safe_metadata_pattern='"[^"]*(secret|token)[^"]*"[[:space:]]*:[[:space:]]*"WECOM_[A-Z0-9_]*(SECRET|TOKEN)([[:space:]]|["]|$)|"[^"]*(secret|token)[^"]*"[[:space:]]*:[[:space:]]*"WECOM_[A-Z0-9_]*(SECRET|TOKEN)[^"]*(source|path|read paths|fallback)'
+  local safe_redacted_prefix_pattern='"[^"]*(sp_no|spNo|template_id|templateId|userid|userId|external_userid|externalUserId|external_user_id)[^"]*Prefixes?"[[:space:]]*:[[:space:]]*"[^"]{1,16}\*\*\*[^"]{0,8}"'
+  local matches
+  matches="$(grep -RInEi "$sensitive_pattern" "$REPORT_DIR" | grep -vE 'HRWDTEST-[0-9]{14}' | grep -vEi "$safe_metadata_pattern" | grep -vEi "$safe_redacted_prefix_pattern" || true)"
+  if [[ -n "$matches" ]]; then
+    printf '%s\n' "$matches"
+    log "[wext1f][FAIL] report-privacy-smoke matched sensitive pattern"
+    STATUS=1
+  else
+    log "[wext1f][PASS] report-privacy-smoke"
+  fi
+}
+
+write_summary_json() {
+  node - "$ROOT_DIR" "$REPORT_DIR" "$PUBLIC_BASE_URL" "$STATUS" <<'NODE' || return 1
+const fs = require("node:fs");
+const path = require("node:path");
+const { execSync } = require("node:child_process");
+
+const [rootDir, reportDir, baseURL, statusInput] = process.argv.slice(2);
+
+function readJson(file) {
+  if (!file || !fs.existsSync(file)) return null;
+  try {
+    return JSON.parse(fs.readFileSync(file, "utf8"));
+  } catch {
+    return null;
+  }
+}
+
+function child(relativePath) {
+  const file = path.join(reportDir, relativePath, "summary.json");
+  const json = readJson(file);
+  return {
+    status: json?.status || json?.conclusion || "MISSING",
+    conclusion: json?.conclusion || null,
+    reportPath: fs.existsSync(file) ? path.resolve(file) : null,
+    mainBlockers: Array.isArray(json?.mainBlockers) ? json.mainBlockers : [],
+    configGaps: Array.isArray(json?.configGaps) ? json.configGaps : [],
+    templateDataGaps: Array.isArray(json?.checks?.templateDataGaps) ? json.checks.templateDataGaps : [],
+    findingCount: Number(json?.findingCount || 0),
+  };
+}
+
+function safeStatus(value) {
+  return String(value || "UNKNOWN").slice(0, 120);
+}
+
+function isPassLike(value) {
+  return /^(PASS|PASS_|PASS_RULE_MATCH|PASS_WITH_ADMIN_FALLBACK|PASS_ARCHIVE_HISTORY_READABLE)/i.test(safeStatus(value));
+}
+
+function isFindingPass(value) {
+  return /^PASS_WITH_FINDINGS$/i.test(safeStatus(value));
+}
+
+function isNoDataGap(report) {
+  const reasons = [
+    ...report.configGaps.map((item) => item?.reason || item?.id),
+    ...report.mainBlockers.map((item) => item?.reason || item?.id),
+    ...report.templateDataGaps.map((item) => item?.reason || item?.id),
+  ].filter(Boolean).join(" ");
+  return /no_wecom_approval_archive_rows|no_existing_approval_in_window|archive_history_data|approval_compensation_window_empty/i.test(reasons);
+}
+
+let commit = "unknown";
+try {
+  commit = execSync("git rev-parse --short HEAD", { cwd: rootDir, encoding: "utf8", stdio: ["ignore", "pipe", "ignore"] }).trim();
+} catch {}
+
+const childReports = {
+  publicReady: {
+    status: Number(statusInput || 0) === 0 ? "PASS" : "FAIL",
+    reportPath: path.resolve(path.join(reportDir, "public-ready.log")),
+  },
+  compensationArchive: child("wecom-approval-compensation-poc"),
+  archiveOffline: child("wecom-approval-archive"),
+  ruleValidator: child("wecom-approval-rule-validator"),
+};
+
+const findings = [];
+let hasBlocker = Number(statusInput || 0) !== 0;
+let hasManualGap = false;
+let hasFinding = false;
+
+for (const [source, report] of Object.entries(childReports)) {
+  const status = safeStatus(report.status);
+  if (isPassLike(status)) {
+    if (isFindingPass(status) || Number(report.findingCount || 0) > 0) {
+      hasFinding = true;
+      findings.push({ source, status, severity: "finding", reportPath: report.reportPath });
+    }
+    continue;
+  }
+  if (/PARTIAL|PAUSED|MISSING|NEEDS/i.test(status) || isNoDataGap(report)) {
+    hasManualGap = true;
+    findings.push({ source, status, severity: "manual_or_data_gap", reportPath: report.reportPath });
+    continue;
+  }
+  hasBlocker = true;
+  findings.push({ source, status, severity: "blocker", reportPath: report.reportPath });
+}
+
+let status = "PASS_FORMAL_OUTING_OVERTIME_READY";
+if (hasBlocker) {
+  status = "SKIP/BLOCKED_BY_EXTERNAL_CONFIG";
+} else if (hasManualGap) {
+  status = "PAUSED_MANUAL_CONFIRMATION";
+} else if (hasFinding) {
+  status = "PASS_WITH_FINDINGS";
+}
+
+const summary = {
+  issueId: "WEXT1F",
+  mode: "wecom-formal-outing-overtime-evidence",
+  status,
+  conclusion: status,
+  baseURL,
+  commit,
+  executionMode: process.env.WEXT1F_EXECUTION_MODE || "unknown",
+  trustedIpBoundary: "Enterprise WeChat API checks must run on the deployment server; local runs are blocked by default",
+  checkedAt: new Date().toISOString(),
+  reportDir: path.resolve(reportDir),
+  accounts: {
+    applicant: "SMOKE002",
+    approver: "SMOKE101",
+    adminObserverEmployeeNo: "HYD000006",
+  },
+  childReports,
+  findingCount: findings.length,
+  findings,
+  manualPrerequisites: [
+    "SMOKE002 has manually submitted formal outing and overtime approval records.",
+    "SMOKE101 has opened detail pages before approving or rejecting.",
+    "WAPPR1 manual gate remains all:A for outing and overtime.",
+  ],
+  manualConfirmation: {
+    required: "SMOKE002_SUBMITTED_SMOKE101_REVIEWED",
+    provided: process.env.WEXT1F_MANUAL_CONFIRMATION === "SMOKE002_SUBMITTED_SMOKE101_REVIEWED",
+  },
+  boundaries: {
+    noWeComWrites: true,
+    noWeComTemplateCreation: true,
+    noWeComApprovalAutoSubmission: true,
+    noRealEmployees: true,
+    onlyAllowedOaWrites: "approval archive persistence via WECOM_ARCHIVE_WRITE=1 + --persist-to-db",
+  },
+  nextQuestions: {
+    wreal1: [
+      "A：1 名真实员工 + 1 名审批人/管理员观察，试用 3 天。",
+      "B：继续测试账号，不引入真实员工。",
+      "C：暂停，先处理指定 finding。",
+    ],
+  },
+  privacy: "status/count/report-path summaries only; no phone/email/token/Secret/EncodingAESKey/full userid/full template_id/full sp_no/form values/attachments/comments/raw payload/XML/ciphertext",
+};
+
+fs.writeFileSync(path.join(reportDir, "summary.json"), `${JSON.stringify(summary, null, 2)}\n`, "utf8");
+console.log(`[wext1f] summary=${path.join(reportDir, "summary.json")}`);
+NODE
+}
+
+finish() {
+  write_summary_json || STATUS=1
+  run_privacy_smoke
+  write_summary_json || STATUS=1
+  log "[wext1f] report=$REPORT_DIR"
+  exit "$STATUS"
+}
+
+trap finish EXIT
+
+if [[ "${WEXT1F_MANUAL_CONFIRMATION:-}" != "$REQUIRED_MANUAL_CONFIRMATION" ]]; then
+  log "[wext1f][PAUSED] missing manual confirmation: set WEXT1F_MANUAL_CONFIRMATION=$REQUIRED_MANUAL_CONFIRMATION only after SMOKE002 submitted formal outing/overtime and SMOKE101 reviewed details."
+  exit 0
+fi
+
+run_ready
+run_formal_compensation_archive
+run_archive_offline
+run_rule_validator
diff --git a/server/src/scripts/check-wecom-approval-archive.js b/server/src/scripts/check-wecom-approval-archive.js
index f444a5a3..8fc8d4dc 100644
--- a/server/src/scripts/check-wecom-approval-archive.js
+++ b/server/src/scripts/check-wecom-approval-archive.js
@@ -9,6 +9,7 @@ const { maskApprovalIdentifier } = require("../services/integrations/wecomApprov
 
 const DEFAULT_REPORT_ROOT = path.resolve(__dirname, "../../../artifacts/qa-flow");
 const DEFAULT_TARGETS = Object.freeze(["leave", "missed_punch"]);
+const SUPPORTED_TARGETS = Object.freeze(["leave", "missed_punch", "outing", "overtime"]);
 const REPORT_PRIVACY =
   "OA DB archive counts/status summaries/masked sp_no/template/userid prefixes only; no phone/email/token/secret/full userid/full template_id/full sp_no/reason text/attachments/comments/raw payload";
 
@@ -24,6 +25,8 @@ Options:
   --report-dir=<path>            报告目录，默认 artifacts/qa-flow/wecom-approval-archive-<timestamp>。
   --template=leave               只检查请假留存。
   --template=missed_punch        只检查补卡留存。
+  --template=outing              只检查外出留存。
+  --template=overtime            只检查加班留存。
   --status=<status>              按规范化状态过滤。
   --limit=<1..100>               最大读取条数，默认 20。
   --help                         显示帮助。
@@ -102,7 +105,7 @@ function parseArgs(argv) {
   }
   const uniqueTemplates = Array.from(new Set(options.templates.filter(Boolean)));
   for (const template of uniqueTemplates) {
-    if (!DEFAULT_TARGETS.includes(template)) {
+    if (!SUPPORTED_TARGETS.includes(template)) {
       throw new Error(`Unsupported --template: ${template}`);
     }
   }
diff --git a/server/src/scripts/check-wecom-approval-compensation-poc.js b/server/src/scripts/check-wecom-approval-compensation-poc.js
index 501373c3..7f38c91d 100644
--- a/server/src/scripts/check-wecom-approval-compensation-poc.js
+++ b/server/src/scripts/check-wecom-approval-compensation-poc.js
@@ -9,6 +9,7 @@ const { getConfig } = require("../config");
 const { createPrismaClient } = require("../prismaClient");
 const {
   APPROVAL_GET_INFO_ENDPOINT,
+  APPROVAL_HR_ATTENDANCE_TARGETS,
   APPROVAL_POC_TARGETS,
   fetchWecomApprovalDetail,
   fetchWecomApprovalInfoList,
@@ -35,7 +36,10 @@ const REPORT_PRIVACY =
 const TEMPLATE_ENV = Object.freeze({
   leave: "WECOM_APPROVAL_LEAVE_TEMPLATE_ID",
   missed_punch: "WECOM_APPROVAL_MISSED_PUNCH_TEMPLATE_ID",
+  outing: "WECOM_APPROVAL_OUTING_TEMPLATE_ID",
+  overtime: "WECOM_APPROVAL_OVERTIME_TEMPLATE_ID",
 });
+const DEFAULT_TARGET_KEYS = Object.freeze(APPROVAL_POC_TARGETS.map((item) => item.key));
 
 function printHelp() {
   process.stdout.write(`Usage: node src/scripts/check-wecom-approval-compensation-poc.js [options]
@@ -51,9 +55,15 @@ Options:
   --end-time=<iso|epoch>         窗口结束时间，默认当前时间。
   --template=leave:<id>          指定请假模板 ID，用于 template_id filter。
   --template=missed_punch:<id>   指定补卡模板 ID，用于 template_id filter。
+  --template=outing:<id>         指定外出模板 ID，用于 WEXT1-FORMAL 正式模板只读证据。
+  --template=overtime:<id>       指定加班模板 ID，用于 WEXT1-FORMAL 正式模板只读证据。
+  --target=<key>                 显式目标，可重复；默认 leave + missed_punch。
+  --targets=<a,b>                显式目标列表；支持 leave, missed_punch, outing, overtime。
   --discover-template-ids        缺模板 ID 时，按申请人只读发现 OA-POC 模板 ID。
   --discover-applicant-employee-no=<no>
                                  自动发现使用的 OA 申请人员工号，默认 SMOKE002。
+  --filter-creator-employee-no=<no>
+                                 拉取模板详情时同时按创建人员工号过滤；正式模板证据必须设置。
   --size=<1..100>                单页审批单号数量，默认 100。
   --persist-to-db                写入 OA 自建留存表；还必须设置 WECOM_ARCHIVE_WRITE=1。
   --help                         显示帮助。
@@ -86,7 +96,26 @@ function safeMessage(value) {
 }
 
 function supportedTargetKeys() {
-  return APPROVAL_POC_TARGETS.map((item) => item.key);
+  return APPROVAL_HR_ATTENDANCE_TARGETS.map((item) => item.key);
+}
+
+function selectedTargetKeys(options) {
+  const rawTargets = Array.isArray(options.targets) && options.targets.length
+    ? options.targets
+    : DEFAULT_TARGET_KEYS;
+  const uniqueTargets = Array.from(new Set(rawTargets.map((item) => toSafeString(item, 40)).filter(Boolean)));
+  const supported = supportedTargetKeys();
+  for (const target of uniqueTargets) {
+    if (!supported.includes(target)) {
+      throw new Error(`unsupported target: ${target}`);
+    }
+  }
+  return uniqueTargets.length ? uniqueTargets : [...DEFAULT_TARGET_KEYS];
+}
+
+function selectedTargetDescriptors(options) {
+  const selected = new Set(selectedTargetKeys(options));
+  return APPROVAL_HR_ATTENDANCE_TARGETS.filter((item) => selected.has(item.key));
 }
 
 function parseTargetPair(value, flagName) {
@@ -114,8 +143,10 @@ function parseArgs(argv) {
     startTime: "",
     endTime: "",
     templates: {},
+    targets: [],
     discoverTemplateIds: false,
     discoverApplicantEmployeeNo: DEFAULT_DISCOVERY_APPLICANT_EMPLOYEE_NO,
+    filterCreatorEmployeeNo: "",
     size: DEFAULT_PAGE_SIZE,
     maxPages: DEFAULT_MAX_PAGES,
     persistToDb: false,
@@ -157,6 +188,16 @@ function parseArgs(argv) {
     } else if (arg.startsWith("--template=")) {
       const pair = parseTargetPair(arg.slice("--template=".length), "--template");
       options.templates[pair.target] = pair.value;
+    } else if (arg === "--target") {
+      options.targets.push(toSafeString(argv[index + 1], 40));
+      index += 1;
+    } else if (arg.startsWith("--target=")) {
+      options.targets.push(toSafeString(arg.slice("--target=".length), 40));
+    } else if (arg === "--targets") {
+      options.targets.push(...toSafeString(argv[index + 1], 200).split(",").map((item) => item.trim()).filter(Boolean));
+      index += 1;
+    } else if (arg.startsWith("--targets=")) {
+      options.targets.push(...toSafeString(arg.slice("--targets=".length), 200).split(",").map((item) => item.trim()).filter(Boolean));
     } else if (arg === "--discover-template-ids") {
       options.discoverTemplateIds = true;
     } else if (arg === "--discover-applicant-employee-no") {
@@ -164,6 +205,11 @@ function parseArgs(argv) {
       index += 1;
     } else if (arg.startsWith("--discover-applicant-employee-no=")) {
       options.discoverApplicantEmployeeNo = toSafeString(arg.slice("--discover-applicant-employee-no=".length), 80);
+    } else if (arg === "--filter-creator-employee-no") {
+      options.filterCreatorEmployeeNo = toSafeString(argv[index + 1], 80);
+      index += 1;
+    } else if (arg.startsWith("--filter-creator-employee-no=")) {
+      options.filterCreatorEmployeeNo = toSafeString(arg.slice("--filter-creator-employee-no=".length), 80);
     } else if (arg === "--size") {
       options.size = Number(argv[index + 1] || options.size);
       index += 1;
@@ -187,6 +233,7 @@ function parseArgs(argv) {
   if (!options.discoverApplicantEmployeeNo) {
     throw new Error("--discover-applicant-employee-no requires non-empty value");
   }
+  options.targets = selectedTargetKeys(options);
   options.size = Math.floor(options.size);
   return options;
 }
@@ -226,7 +273,7 @@ function envValue(env, name) {
 }
 
 function buildInputs(options, env = process.env) {
-  return APPROVAL_POC_TARGETS.map((target) => {
+  return selectedTargetDescriptors(options).map((target) => {
     const cliTemplateId = toSafeString(options.templates[target.key], 500);
     const envTemplateId = envValue(env, TEMPLATE_ENV[target.key]);
     const templateId = cliTemplateId || envTemplateId;
@@ -327,6 +374,8 @@ function inferTargetFromApprovalDetail(detailResult) {
   ].filter(Boolean).join(" ");
   if (/补卡|missed[_ -]?punch/i.test(text)) return "missed_punch";
   if (/请假|leave/i.test(text)) return "leave";
+  if (/外出|outing|outgoing/i.test(text)) return "outing";
+  if (/加班|overtime/i.test(text)) return "overtime";
   return null;
 }
 
@@ -615,11 +664,98 @@ function buildDryRunTargetCheck(input) {
   };
 }
 
+function buildCreatorFilterNotReadyTargetCheck(input, creatorFilterCheck) {
+  return {
+    id: "approval_compensation_pull",
+    target: input.key,
+    label: input.label,
+    endpoint: `POST ${APPROVAL_GET_INFO_ENDPOINT}`,
+    source: "WECOM_AGENT_SECRET",
+    status: "NEEDS_CONFIG",
+    reason: creatorFilterCheck?.reason || "creator_filter_not_ready",
+    missingEnv: creatorFilterCheck?.missing || [],
+    templateIdPrefix: input.templateIdPrefix,
+    configuredFrom: input.templateSource,
+    creatorFilter: {
+      requested: true,
+      status: creatorFilterCheck?.status || "NEEDS_CONFIG",
+      employeeNo: creatorFilterCheck?.employeeNo || null,
+      userIdPrefix: creatorFilterCheck?.userIdPrefix || null,
+    },
+    pageCount: 0,
+    spNoCount: 0,
+    details: [],
+  };
+}
+
+function buildSkippedCreatorFilterCheck(reason) {
+  return {
+    id: "approval_creator_filter",
+    status: "SKIPPED",
+    reason,
+    employeeNo: null,
+    applicant: null,
+    userIdPrefix: null,
+  };
+}
+
+async function buildCreatorFilterCheck({ options, queryApplicantActorImpl }) {
+  const employeeNo = toSafeString(options.filterCreatorEmployeeNo, 80);
+  if (!employeeNo) return buildSkippedCreatorFilterCheck("creator_filter_not_requested");
+
+  let applicant;
+  try {
+    applicant = await queryApplicantActorImpl(employeeNo);
+  } catch (error) {
+    return {
+      id: "approval_creator_filter",
+      status: "NEEDS_CONFIG",
+      reason: "creator_filter_lookup_failed",
+      message: safeMessage(error?.message || error),
+      employeeNo,
+      applicant: null,
+      userIdPrefix: null,
+      missing: ["active_employee_wecom_binding"],
+    };
+  }
+
+  const gaps = [];
+  if (!applicant) gaps.push("creator_employee_missing");
+  if (applicant && !isOaActive(applicant)) gaps.push("creator_employee_inactive");
+  if (applicant && !applicant.wecomExternalUserId) gaps.push("creator_employee_missing_wecom_binding");
+  if (gaps.length) {
+    return {
+      id: "approval_creator_filter",
+      status: "NEEDS_CONFIG",
+      reason: "creator_filter_not_ready",
+      employeeNo,
+      applicant: sanitizeDiscoveryActor(applicant),
+      userIdPrefix: null,
+      gaps,
+      missing: ["active_employee_wecom_binding"],
+    };
+  }
+
+  return attachHiddenValue({
+    id: "approval_creator_filter",
+    status: "OK",
+    reason: null,
+    employeeNo,
+    applicant: sanitizeDiscoveryActor(applicant),
+    userIdPrefix: maskApprovalIdentifier(applicant.wecomExternalUserId),
+    filter: {
+      key: "creator",
+      useridPrefix: maskApprovalIdentifier(applicant.wecomExternalUserId),
+    },
+  }, "userid", applicant.wecomExternalUserId);
+}
+
 async function runLiveTarget({
   config,
   input,
   timeWindow,
   options,
+  creatorFilterCheck,
   fetchApprovalInfoListImpl,
   fetchApprovalDetailImpl,
   archivePersistence,
@@ -632,13 +768,17 @@ async function runLiveTarget({
 
   do {
     pageIndex += 1;
+    const filters = [{ key: "template_id", value: input.templateId }];
+    if (creatorFilterCheck?.status === "OK" && creatorFilterCheck.userid) {
+      filters.push({ key: "creator", value: creatorFilterCheck.userid });
+    }
     const result = await fetchApprovalInfoListImpl({
       config,
       startTime: timeWindow.startTime,
       endTime: timeWindow.endTime,
       size: options.size,
       nextCursor: cursor,
-      filters: [{ key: "template_id", value: input.templateId }],
+      filters,
     });
     pages.push({
       page: pageIndex,
@@ -672,6 +812,10 @@ async function runLiveTarget({
       message: blocked.message ? safeMessage(blocked.message) : null,
       templateIdPrefix: input.templateIdPrefix,
       configuredFrom: input.templateSource,
+      creatorFilter: creatorFilterCheck?.status === "OK" ? {
+        employeeNo: creatorFilterCheck.employeeNo,
+        userIdPrefix: creatorFilterCheck.userIdPrefix,
+      } : null,
       pageCount: pages.length,
       spNoCount: 0,
       pages,
@@ -727,6 +871,10 @@ async function runLiveTarget({
     reason: detailBlocked ? detailBlocked.reason : null,
     templateIdPrefix: input.templateIdPrefix,
     configuredFrom: input.templateSource,
+    creatorFilter: creatorFilterCheck?.status === "OK" ? {
+      employeeNo: creatorFilterCheck.employeeNo,
+      userIdPrefix: creatorFilterCheck.userIdPrefix,
+    } : null,
     pageCount: pages.length,
     spNoCount: uniqueSpNos.length,
     pages,
@@ -740,6 +888,7 @@ async function runTargetChecks({
   inputs,
   timeWindow,
   options,
+  creatorFilterCheck,
   fetchApprovalInfoListImpl,
   fetchApprovalDetailImpl,
   archivePersistence,
@@ -748,6 +897,8 @@ async function runTargetChecks({
   for (const input of inputs) {
     if (!input.templateId) {
       targetChecks.push(buildMissingTemplateCheck(input));
+    } else if (options.filterCreatorEmployeeNo && creatorFilterCheck?.status !== "OK") {
+      targetChecks.push(buildCreatorFilterNotReadyTargetCheck(input, creatorFilterCheck));
     } else if (!options.live) {
       targetChecks.push(buildDryRunTargetCheck(input));
     } else {
@@ -756,6 +907,7 @@ async function runTargetChecks({
         input,
         timeWindow,
         options,
+        creatorFilterCheck,
         fetchApprovalInfoListImpl,
         fetchApprovalDetailImpl,
         archivePersistence,
@@ -956,7 +1108,7 @@ function collectConfigGaps(targetChecks, verificationGaps, evidenceCheck, compar
     gaps.push({
       id: "approval_compensation_window_empty",
       status: "NEEDS_CONFIG",
-      reason: "no_existing_oa_poc_approval_in_window",
+      reason: "no_existing_approval_in_window_for_selected_targets",
     });
   }
   return gaps;
@@ -1148,6 +1300,10 @@ async function executeCompensationPoc(options, deps = {}) {
     fetchApprovalDetailImpl: deps.fetchApprovalDetailImpl || fetchWecomApprovalDetail,
     queryApplicantActorImpl: deps.queryApplicantActorImpl || queryApplicantActorRowReadOnly,
   });
+  const creatorFilterCheck = await buildCreatorFilterCheck({
+    options,
+    queryApplicantActorImpl: deps.queryApplicantActorImpl || queryApplicantActorRowReadOnly,
+  });
   const resolvedInputs = applyDiscoveredTemplateIds(inputs, discoveryCheck);
   if (archivePersistenceCheck.status === "READY") {
     try {
@@ -1183,6 +1339,7 @@ async function executeCompensationPoc(options, deps = {}) {
     inputs: resolvedInputs,
     timeWindow,
     options,
+    creatorFilterCheck,
     fetchApprovalInfoListImpl: deps.fetchApprovalInfoListImpl || fetchWecomApprovalInfoList,
     fetchApprovalDetailImpl: deps.fetchApprovalDetailImpl || fetchWecomApprovalDetail,
     archivePersistence,
@@ -1245,7 +1402,7 @@ async function executeCompensationPoc(options, deps = {}) {
     privacy: REPORT_PRIVACY,
     scope: {
       issue: "WACP3",
-      targets: APPROVAL_POC_TARGETS.map((item) => item.key),
+      targets: selectedTargetKeys(options),
       writeBehavior: options.persistToDb
         ? "WeCom read-only; OA archive DB persistence requires --persist-to-db + WECOM_ARCHIVE_WRITE=1"
         : "read-only only; no OA writes, no WeCom writes, no template creation, no approval submission",
@@ -1265,10 +1422,15 @@ async function executeCompensationPoc(options, deps = {}) {
       pageSize: options.size,
       maxPages: options.maxPages,
       discoveryApplicantEmployeeNo: options.discoverTemplateIds ? options.discoverApplicantEmployeeNo : null,
+      filterCreatorEmployeeNo: options.filterCreatorEmployeeNo || null,
     },
     checks: {
       secretMatrix: secretCheck,
       templateDiscovery: sanitizeDiscoveryCheck(discoveryCheck),
+      creatorFilter: {
+        ...creatorFilterCheck,
+        userid: undefined,
+      },
       targets: targetChecks,
       callbackEvidence: {
         ...evidenceCheck,
diff --git a/server/src/services/integrations/wecomApproval.js b/server/src/services/integrations/wecomApproval.js
index d1989211..e414ce3f 100644
--- a/server/src/services/integrations/wecomApproval.js
+++ b/server/src/services/integrations/wecomApproval.js
@@ -9,6 +9,11 @@ const APPROVAL_POC_TARGETS = Object.freeze([
   { key: "leave", label: "请假" },
   { key: "missed_punch", label: "补卡" },
 ]);
+const APPROVAL_HR_ATTENDANCE_TARGETS = Object.freeze([
+  ...APPROVAL_POC_TARGETS,
+  { key: "outing", label: "外出" },
+  { key: "overtime", label: "加班" },
+]);
 
 class WecomApprovalPocError extends Error {
   constructor(code, message, details = {}) {
@@ -650,6 +655,7 @@ async function fetchWecomApprovalApplyEvent({
 }
 
 module.exports = {
+  APPROVAL_HR_ATTENDANCE_TARGETS,
   APPROVAL_POC_TARGETS,
   APPROVAL_APPLY_EVENT_ENDPOINT,
   APPROVAL_CREATE_TEMPLATE_ENDPOINT,
diff --git a/server/tests/unit/scripts/check-wecom-approval-archive.test.js b/server/tests/unit/scripts/check-wecom-approval-archive.test.js
index fc9e4d6d..cfdac472 100644
--- a/server/tests/unit/scripts/check-wecom-approval-archive.test.js
+++ b/server/tests/unit/scripts/check-wecom-approval-archive.test.js
@@ -77,7 +77,8 @@ describe("check-wecom-approval-archive", () => {
       status: "approved",
       limit: 50,
     });
-    expect(() => parseArgs(["--template=outing"])).toThrow(/Unsupported --template/);
+    expect(parseArgs(["--template=outing", "--template=overtime"]).templates).toEqual(["outing", "overtime"]);
+    expect(() => parseArgs(["--template=unknown"])).toThrow(/Unsupported --template/);
     expect(() => parseArgs(["--limit=101"])).toThrow(/--limit/);
   });
 
diff --git a/server/tests/unit/scripts/check-wecom-approval-compensation-poc.test.js b/server/tests/unit/scripts/check-wecom-approval-compensation-poc.test.js
index 236d14e4..ae1aeee9 100644
--- a/server/tests/unit/scripts/check-wecom-approval-compensation-poc.test.js
+++ b/server/tests/unit/scripts/check-wecom-approval-compensation-poc.test.js
@@ -89,8 +89,11 @@ describe("check-wecom-approval-compensation-poc", () => {
       "--size=50",
       "--template=leave:tpl-leave",
       "--template=missed_punch:tpl-missed",
+      "--target=leave",
+      "--targets=missed_punch",
       "--discover-template-ids",
       "--discover-applicant-employee-no=SMOKE002",
+      "--filter-creator-employee-no=SMOKE002",
     ]);
 
     expect(options.live).toBe(true);
@@ -98,12 +101,16 @@ describe("check-wecom-approval-compensation-poc", () => {
     expect(options.size).toBe(50);
     expect(options.discoverTemplateIds).toBe(true);
     expect(options.discoverApplicantEmployeeNo).toBe("SMOKE002");
+    expect(options.filterCreatorEmployeeNo).toBe("SMOKE002");
     expect(options.persistToDb).toBe(false);
+    expect(options.targets).toEqual(["leave", "missed_punch"]);
     expect(options.templates).toEqual({
       leave: "tpl-leave",
       missed_punch: "tpl-missed",
     });
-    expect(() => parseArgs(["--template=outing:tpl"])).toThrow(/unsupported target/);
+    expect(parseArgs(["--target=outing", "--targets=overtime"]).targets).toEqual(["outing", "overtime"]);
+    expect(parseArgs(["--template=outing:tpl"]).templates).toMatchObject({ outing: "tpl" });
+    expect(() => parseArgs(["--target=unknown"])).toThrow(/unsupported target/);
     expect(() => parseArgs(["--size=101"])).toThrow(/--size/);
     expect(parseArgs(["--persist-to-db"]).persistToDb).toBe(true);
   });
@@ -133,6 +140,29 @@ describe("check-wecom-approval-compensation-poc", () => {
     });
   });
 
+  it("builds explicit WEXT1-FORMAL outing and overtime inputs without changing defaults", () => {
+    const defaultInputs = buildInputs(baseOptions({ templates: {} }), {});
+    expect(defaultInputs.map((item) => item.key)).toEqual(["leave", "missed_punch"]);
+
+    const inputs = buildInputs(baseOptions({
+      targets: ["outing", "overtime"],
+      templates: { outing: "cli-outing-template" },
+    }), {
+      WECOM_APPROVAL_OUTING_TEMPLATE_ID: "env-outing-template",
+      WECOM_APPROVAL_OVERTIME_TEMPLATE_ID: "env-overtime-template",
+    });
+
+    expect(inputs.map((item) => item.key)).toEqual(["outing", "overtime"]);
+    expect(inputs.find((item) => item.key === "outing")).toMatchObject({
+      templateSource: "cli",
+      templateIdPrefix: maskApprovalIdentifier("cli-outing-template"),
+    });
+    expect(inputs.find((item) => item.key === "overtime")).toMatchObject({
+      templateSource: "WECOM_APPROVAL_OVERTIME_TEMPLATE_ID",
+      templateIdPrefix: maskApprovalIdentifier("env-overtime-template"),
+    });
+  });
+
   it("documents WECOM_AGENT_SECRET as the only Secret source", () => {
     const check = buildSecretCheck(baseConfig());
     expect(check).toMatchObject({
@@ -227,6 +257,8 @@ describe("check-wecom-approval-compensation-poc", () => {
   it("auto-discovers missing template IDs by applicant creator without leaking full identifiers", async () => {
     const discoveredLeaveTemplateId = "tpllv-auto-discovered-leave-sensitive";
     const discoveredMissedTemplateId = "tplmp-auto-discovered-missed-sensitive";
+    const discoveredOutingTemplateId = "tplout-auto-discovered-outing-sensitive";
+    const discoveredOvertimeTemplateId = "tplot-auto-discovered-overtime-sensitive";
     const fetchApprovalInfoListImpl = vi.fn(async ({ filters }) => {
       if (filters[0].key === "creator") {
         return {
@@ -234,10 +266,15 @@ describe("check-wecom-approval-compensation-poc", () => {
           tokenSource: "WECOM_AGENT_SECRET",
           response: {
             errcode: 0,
-            spNoCount: 2,
-            spNoPrefixes: [maskApprovalIdentifier("sp-discovery-leave"), maskApprovalIdentifier("sp-discovery-missed")],
+            spNoCount: 4,
+            spNoPrefixes: [
+              maskApprovalIdentifier("sp-discovery-leave"),
+              maskApprovalIdentifier("sp-discovery-missed"),
+              maskApprovalIdentifier("sp-discovery-outing"),
+              maskApprovalIdentifier("sp-discovery-overtime"),
+            ],
           },
-          spNos: ["sp-discovery-leave", "sp-discovery-missed"],
+          spNos: ["sp-discovery-leave", "sp-discovery-missed", "sp-discovery-outing", "sp-discovery-overtime"],
           nextCursor: "",
         };
       }
@@ -274,13 +311,37 @@ describe("check-wecom-approval-compensation-poc", () => {
           response: { errcode: 0, spStatus: "2" },
         };
       }
+      if (spNo === "sp-discovery-outing") {
+        return {
+          ok: true,
+          tokenSource: "WECOM_AGENT_SECRET",
+          templateId: discoveredOutingTemplateId,
+          spName: "外出",
+          summary: detailSummary(spNo, { templateIdPrefix: maskApprovalIdentifier(discoveredOutingTemplateId) }),
+          response: { errcode: 0, spStatus: "2" },
+        };
+      }
+      if (spNo === "sp-discovery-overtime") {
+        return {
+          ok: true,
+          tokenSource: "WECOM_AGENT_SECRET",
+          templateId: discoveredOvertimeTemplateId,
+          spName: "加班",
+          summary: detailSummary(spNo, { templateIdPrefix: maskApprovalIdentifier(discoveredOvertimeTemplateId) }),
+          response: { errcode: 0, spStatus: "2" },
+        };
+      }
       return {
         ok: true,
         tokenSource: "WECOM_AGENT_SECRET",
         summary: detailSummary(spNo, {
           templateIdPrefix: spNo.startsWith(discoveredLeaveTemplateId)
             ? maskApprovalIdentifier(discoveredLeaveTemplateId)
-            : maskApprovalIdentifier(discoveredMissedTemplateId),
+            : spNo.startsWith(discoveredMissedTemplateId)
+              ? maskApprovalIdentifier(discoveredMissedTemplateId)
+              : spNo.startsWith(discoveredOutingTemplateId)
+                ? maskApprovalIdentifier(discoveredOutingTemplateId)
+                : maskApprovalIdentifier(discoveredOvertimeTemplateId),
         }),
         response: { errcode: 0, spStatus: "2" },
       };
@@ -288,6 +349,7 @@ describe("check-wecom-approval-compensation-poc", () => {
 
     const report = await executeCompensationPoc(baseOptions({
       templates: {},
+      targets: ["leave", "missed_punch", "outing", "overtime"],
       discoverTemplateIds: true,
       discoverApplicantEmployeeNo: "SMOKE002",
     }), {
@@ -315,6 +377,18 @@ describe("check-wecom-approval-compensation-poc", () => {
           statusLabel: "已通过",
           spNoPrefix: maskApprovalIdentifier(`${discoveredMissedTemplateId}-sp`),
         },
+        {
+          callbackKind: "approval_status_change",
+          spStatus: "2",
+          statusLabel: "已通过",
+          spNoPrefix: maskApprovalIdentifier(`${discoveredOutingTemplateId}-sp`),
+        },
+        {
+          callbackKind: "approval_status_change",
+          spStatus: "2",
+          statusLabel: "已通过",
+          spNoPrefix: maskApprovalIdentifier(`${discoveredOvertimeTemplateId}-sp`),
+        },
       ]),
     });
 
@@ -328,13 +402,23 @@ describe("check-wecom-approval-compensation-poc", () => {
     expect(fetchApprovalInfoListImpl.mock.calls[2][0].filters).toEqual([
       { key: "template_id", value: discoveredMissedTemplateId },
     ]);
+    expect(fetchApprovalInfoListImpl.mock.calls[3][0].filters).toEqual([
+      { key: "template_id", value: discoveredOutingTemplateId },
+    ]);
+    expect(fetchApprovalInfoListImpl.mock.calls[4][0].filters).toEqual([
+      { key: "template_id", value: discoveredOvertimeTemplateId },
+    ]);
     const serialized = JSON.stringify(report);
     expect(serialized).not.toContain(discoveredLeaveTemplateId);
     expect(serialized).not.toContain(discoveredMissedTemplateId);
+    expect(serialized).not.toContain(discoveredOutingTemplateId);
+    expect(serialized).not.toContain(discoveredOvertimeTemplateId);
     expect(serialized).not.toContain(SENSITIVE_USER_ID);
     expect(report.checks.templateDiscovery.discoveredTemplatePrefixes).toEqual({
       leave: maskApprovalIdentifier(discoveredLeaveTemplateId),
       missed_punch: maskApprovalIdentifier(discoveredMissedTemplateId),
+      outing: maskApprovalIdentifier(discoveredOutingTemplateId),
+      overtime: maskApprovalIdentifier(discoveredOvertimeTemplateId),
     });
   });
 
@@ -402,6 +486,107 @@ describe("check-wecom-approval-compensation-poc", () => {
     expect(report.checks.archivePersistence.status).toBe("SKIPPED");
   });
 
+  it("filters WEXT1-FORMAL template pulls by creator employee to avoid broad formal-template reads", async () => {
+    const creatorUserId = "formal_creator_userid_should_not_leak";
+    const outingTemplateId = "outing-template-sensitive";
+    const overtimeTemplateId = "overtime-template-sensitive";
+    const fetchApprovalInfoListImpl = vi.fn(async ({ filters }) => {
+      expect(filters).toHaveLength(2);
+      expect(filters[0].key).toBe("template_id");
+      expect(filters[1]).toEqual({ key: "creator", value: creatorUserId });
+      return {
+        ok: true,
+        tokenSource: "WECOM_AGENT_SECRET",
+        response: { errcode: 0, spNoCount: 1, spNoPrefixes: [maskApprovalIdentifier(`${filters[0].value}-sp`)] },
+        spNos: [`${filters[0].value}-sp`],
+        nextCursor: "",
+      };
+    });
+    const fetchApprovalDetailImpl = vi.fn(async ({ spNo }) => ({
+      ok: true,
+      tokenSource: "WECOM_AGENT_SECRET",
+      summary: detailSummary(spNo, { spStatus: "2" }),
+      response: { errcode: 0, spStatus: "2" },
+    }));
+
+    const report = await executeCompensationPoc(baseOptions({
+      targets: ["outing", "overtime"],
+      templates: {
+        outing: outingTemplateId,
+        overtime: overtimeTemplateId,
+      },
+      filterCreatorEmployeeNo: "SMOKE002",
+    }), {
+      config: baseConfig(),
+      env: {},
+      generatedAt: "2026-05-20T00:00:00.000Z",
+      queryApplicantActorImpl: vi.fn(async () => ({
+        employeeNo: "SMOKE002",
+        status: "active",
+        archived: false,
+        wecomExternalUserId: creatorUserId,
+      })),
+      fetchApprovalInfoListImpl,
+      fetchApprovalDetailImpl,
+      readCallbackEvidenceImpl: vi.fn(async () => [
+        {
+          callbackKind: "approval_status_change",
+          spStatus: "2",
+          statusLabel: "已通过",
+          spNoPrefix: maskApprovalIdentifier(`${outingTemplateId}-sp`),
+        },
+        {
+          callbackKind: "approval_status_change",
+          spStatus: "2",
+          statusLabel: "已通过",
+          spNoPrefix: maskApprovalIdentifier(`${overtimeTemplateId}-sp`),
+        },
+      ]),
+    });
+
+    expect(report.status).toBe("PASS_COMPENSATION_POC");
+    expect(report.checks.creatorFilter).toMatchObject({
+      status: "OK",
+      employeeNo: "SMOKE002",
+      userIdPrefix: maskApprovalIdentifier(creatorUserId),
+    });
+    expect(fetchApprovalInfoListImpl).toHaveBeenCalledTimes(2);
+    const serialized = JSON.stringify(report);
+    expect(serialized).not.toContain(creatorUserId);
+    expect(serialized).not.toContain(outingTemplateId);
+    expect(serialized).not.toContain(overtimeTemplateId);
+  });
+
+  it("does not pull formal templates when the requested creator filter cannot be resolved", async () => {
+    const fetchApprovalInfoListImpl = vi.fn();
+    const report = await executeCompensationPoc(baseOptions({
+      targets: ["outing"],
+      templates: { outing: "formal-outing-template-sensitive" },
+      filterCreatorEmployeeNo: "SMOKE002",
+    }), {
+      config: baseConfig(),
+      env: {},
+      generatedAt: "2026-05-20T00:00:00.000Z",
+      queryApplicantActorImpl: vi.fn(async () => ({
+        employeeNo: "SMOKE002",
+        status: "active",
+        archived: false,
+        wecomExternalUserId: "",
+      })),
+      fetchApprovalInfoListImpl,
+      fetchApprovalDetailImpl: vi.fn(),
+      readCallbackEvidenceImpl: vi.fn(async () => []),
+    });
+
+    expect(report.status).toBe("PARTIAL_COMPENSATION_NEEDS_CONFIG");
+    expect(fetchApprovalInfoListImpl).not.toHaveBeenCalled();
+    expect(report.checks.targets[0]).toMatchObject({
+      target: "outing",
+      status: "NEEDS_CONFIG",
+      reason: "creator_filter_not_ready",
+    });
+  });
+
   it("requires both --persist-to-db and WECOM_ARCHIVE_WRITE=1 before OA archive writes", async () => {
     const report = await executeCompensationPoc(baseOptions({
       live: true,